[House Hearing, 112 Congress]
[From the U.S. Government Printing Office]
EXAMINING THE HOMELAND SECURITY IMPACT OF THE OBAMA ADMINISTRATION'S
CYBERSECURITY PROPOSAL
=======================================================================
HEARING
before the
SUBCOMMITTEE ON CYBERSECURITY,
INFRASTRUCTURE PROTECTION,
AND SECURITY TECHNOLOGIES
of the
COMMITTEE ON HOMELAND SECURITY
HOUSE OF REPRESENTATIVES
ONE HUNDRED TWELFTH CONGRESS
FIRST SESSION
__________
JUNE 24, 2011
__________
Serial No. 112-33
__________
Printed for the use of the Committee on Homeland Security
[GRAPHIC] [TIFF OMITTED] TONGRESS.#13
Available via the World Wide Web: http://www.gpo.gov/fdsys/
__________
U.S. GOVERNMENT PRINTING OFFICE
72-253 WASHINGTON : 2012
---------------------------------------------------------------------
For sale by the Superintendent of Documents, U.S. Government Printing
Office Internet: bookstore.gpo.gov Phone: toll free (866) 512-1800; DC
area (202) 512-1800 Fax: (202) 512-2104 Mail: Stop IDCC, Washington, DC
20402-0001
COMMITTEE ON HOMELAND SECURITY
Peter T. King, New York, Chairman
Lamar Smith, Texas Bennie G. Thompson, Mississippi
Daniel E. Lungren, California Loretta Sanchez, California
Mike Rogers, Alabama Sheila Jackson Lee, Texas
Michael T. McCaul, Texas Henry Cuellar, Texas
Gus M. Bilirakis, Florida Yvette D. Clarke, New York
Paul C. Broun, Georgia Laura Richardson, California
Candice S. Miller, Michigan Danny K. Davis, Illinois
Tim Walberg, Michigan Brian Higgins, New York
Chip Cravaack, Minnesota Jackie Speier, California
Joe Walsh, Illinois Cedric L. Richmond, Louisiana
Patrick Meehan, Pennsylvania Hansen Clarke, Michigan
Ben Quayle, Arizona William R. Keating, Massachusetts
Scott Rigell, Virginia Kathleen C. Hochul, New York
Billy Long, Missouri Vacancy
Jeff Duncan, South Carolina
Tom Marino, Pennsylvania
Blake Farenthold, Texas
Mo Brooks, Alabama
Michael J. Russell, Staff Director/Chief Counsel
Kerry Ann Watkins, Senior Policy Director
Michael S. Twinchek, Chief Clerk
I. Lanier Avant, Minority Staff Director
------
SUBCOMMITTEE ON CYBERSECURITY, INFRASTRUCTURE PROTECTION, AND SECURITY
TECHNOLOGIES
Daniel E. Lungren, California, Chairman
Michael T. McCaul, Texas Yvette D. Clarke, New York
Tim Walberg, Michigan, Vice Chair Laura Richardson, California
Patrick Meehan, Pennsylvania Cedric L. Richmond, Louisiana
Billy Long, Missouri William R. Keating, Massachusetts
Tom Marino, Pennsylvania Bennie G. Thompson, Mississippi
Peter T. King, New York (Ex (Ex Officio)
Officio)
Coley C. O'Brien, Staff Director
Alan Carroll, Subcommittee Clerk
Chris Schepis, Minority Senior Professional Staff Member
C O N T E N T S
----------
Page
Statements
The Honorable Daniel E. Lungren, a Representative in Congress
From the State of California, and Chairman, Subcommittee on
Cybersecurity, Infrastructure Protection, and Security
Technologies................................................... 1
The Honorable Yvette D. Clark, a Representative in Congress From
the State of New York, and Ranking Member, Subcommittee on
Cybersecurity, Infrastructure Protection, and Security
Technologies:
Oral Statement................................................. 3
Prepared Statement............................................. 4
The Honorable Bennie G. Thompson, a Representative in Congress
From the State of Mississippi, and Ranking Member, Committee on
Homeland Security:
Oral Statement................................................. 4
Prepared Statement............................................. 5
Witnesses
Ms. Melissa E. Hathaway, President, Hathaway Global Strategies,
LLC:
Oral Statement................................................. 8
Prepared Statement............................................. 10
Mr. Gregory E. Shannon, Chief Scientist for Computer Emergency
Readiness Team (Cert), Software Engineering Institute, Carnegie
Mellon University:
Oral Statement................................................. 17
Prepared Statement............................................. 18
Mr. Leigh Williams, President, BITS, The Financial Services
Roundtable:
Oral Statement................................................. 21
Prepared Statement............................................. 23
Mr. Larry Clinton, President, Internet Security Alliance:
Oral Statement................................................. 28
Prepared Statement............................................. 30
For the Record
The Honorable Daniel E. Lungren, a Representative in Congress
From the State of California, and Chairman, Subcommittee on
Cybersecurity, Infrastructure Protection, and Security
Technologies:
Statement of the American Chemistry Council.................... 6
Appendix
Questions From Chairman Daniel E. Lungren for Melissa Hathaway... 57
Questions From Chairman Daniel E. Lungren for Gregory E. Shannon. 61
Questions From Chairman Daniel E. Lungren for Leigh Williams..... 64
Questions From Chairman Daniel E. Lungren for Larry Clinton...... 65
EXAMINING THE HOMELAND SECURITY IMPACT OF THE OBAMA ADMINISTRATION'S
CYBERSECURITY PROPOSAL
----------
Friday, June 24, 2011
U.S. House of Representatives,
Committee on Homeland Security,
Subcommittee on Cybersecurity, Infrastructure Protection,
and Security Technologies,
Washington, DC.
The subcommittee met, pursuant to call, at 10:05 a.m., in
Room 311, Cannon House Office Building, Hon. Daniel E. Lungren
[Chairman of the subcommittee] presiding.
Present: Representatives Lungren, McCaul, Walberg, Long,
Marino, Clarke, Richardson, Richmond, and Keating.
Mr. Lungren. With the concurrence of the Ranking Member of
the full committee, the Subcommittee on Cybersecurity,
Infrastructure Protection, and Security Technology will come to
order.
The subcommittee is meeting today to examine the homeland
security impact of the administration's cybersecurity proposal.
I would just say at the outset, we have a vote, I guess a
single vote, scheduled at about 10:15, so we will have to go
over there and then come back. We are going to try and get our
opening statements in so that we can proceed directly with our
witnesses as soon as we get back from the vote.
I recognize myself for an opening statement.
We are meeting today to examine the impact of the
administration's cybersecurity proposal on the Department of
Homeland Security. The proposal touches on a number of issues,
such as increasing the penalty for hacking, putting in place a
comprehensive regime around the issue of large-scale breaches
of personally identifiable information, regulating the
cybersecurity of the private-sector critical infrastructure
owners and operators, and providing needed clarity on the
cybersecurity mission of the Department of Homeland Security.
While I may differ on certain elements of their proposal, I
am pleased the administration has provided thoughtful inputs to
Congress to help us craft an effective National cybersecurity
policy. That being said, I believe this proposal is not the end
of our effort but the beginning of a much-needed debate on how
we, as a Nation, will address these dynamic cybersecurity
threats in the future.
With the growing number of computer network cyber
intrusions and attacks being reported in the media, the need
for strengthening cybersecurity is obviously more evident every
day. The status quo is not acceptable. The internet and our
digital society provide our adversaries multiple attack
avenues.
We must continue to innovate and to build a culture of
cybersecurity. We must also find a way to incentivize critical
infrastructure owners and operators to build security into
their business model. Although our Nation faces a difficult
fiscal environment, securing our critical infrastructure assets
cannot be ignored.
We must be creative and develop ways to improve the return
on our security investments. Developing the right liability
safe harbors for growing a more robust and mature cyber
insurance market won't happen by itself, particularly in this
economic downturn. We must tap the talent of the private sector
to develop the appropriate ways and means to improve the
cybersecurity economic equation. The cost if we don't secure
our critical infrastructure and our business networks and data
will be far greater.
I thank all of our witnesses for their appearances today.
This is the third in our series of hearings on the cyber threat
to critical infrastructure.
The administration's proposal outlines their cybersecurity
vision, and I, frankly, thank them for it. It will help inform
our efforts to develop legislation to better secure our
critical infrastructure and Government networks. I am eager to
hear how the proposed language would impact those in the
private sector, how it would increase the authority of the
Department of Homeland Security, and how it positions the
Department of Homeland Security to be the focal point of our
cybersecurity in the civilian government.
I believe the Department is the appropriate place within
the Government to take responsibility for our cybersecurity
operations and establish policies and priorities for protecting
our civilian departments and agencies. I think having the
Government lead by example is critically important.
As Chairman of the House Committee on Administration, I
have the cybersecurity responsibility for the House of
Representatives. I take that responsibility very seriously and
am proud of the job that the CIO and his team have done. Having
DHS lead by example is critically important, as I mentioned.
We are going to hear from Dr. Greg Shannon this morning
about the future of incident response operations. Carnegie
Mellon CERT has long been recognized for its excellence in
computer emergency response, and I am hopeful that their
experience will help DHS build a world-class computer instant
response capability.
I am also honored to have Melissa Hathaway, the former
cybersecurity advisor to both President Bush and President
Obama, here to discuss the administration's proposal. She was
the director of President Bush's Comprehensive National
Cybersecurity Initiative. Additionally, as primary author of
this administration's 60-Day Cyberspace Policy Review, she is
in a unique position to share with us her expertise and
perspective on improving the overall cybersecurity enterprise
across Government--in particular, how the Government can best
interact with private-sector critical infrastructure owners and
operators.
I applaud the administration for coming forward with a
proposal. They have, I think, some of the answers. I don't
think they believe, nor do I believe, that any one of us has
all of the answers. But, together, we can certainly forge ahead
to improve from where we are now.
As I mentioned, we have had a vote called. Interesting, we
have a little TV screen up here which shows what is going on,
but they have managed to put it in mirror fashion so it is
reverse of what it says. But I do believe that means we have 11
minutes left. Somebody has invaded our little system here.
But I would like to recognize the Ranking Member from New
York, Ms. Clarke, for her opening statement.
Ms. Clarke. Thank you very much, Mr. Chairman, Ranking
Member Thompson, my colleagues, and to the panelists this
morning.
We live in a world where it seems that everything relies on
computers and the internet. The effective functioning of our
critical infrastructure from airports, financial systems, to
water systems, factories, the electric grid is highly dependent
on computer-based systems called control systems that are used
to monitor and control sensitive processes and physical
functions.
The danger of both unintentional and intentional cyber
attack is real. The potential consequences for an attack on
control systems vary widely, from the introduction of raw
sewage into potable water systems to the catastrophic failure
of critical electrical generators due to the change of a single
line of code in the critical system.
We have come to recognize that public-private partnerships
are a key component of securing our Nation's computer-reliant
critical infrastructure. Private-sector involvement is crucial,
as it collectively owns the vast majority of the Nation's cyber
infrastructure and is responsible for protecting its networks
and systems from the growing threat of a cyber attack.
Enhancing the public-private partnerships by developing an
improved value proposition and implementing better incentives,
among other measures, will be essential to encouraging greater
private-sector involvement.
Control systems are not the only computers subject to
attack. Every day, thousands of attacks are launched against
Federal and private networks by hackers, terrorist groups,
nation-states attempting to access classified and unclassified
information. The infiltration by foreign nationals of Federal
Government networks is one of the most pressing issues
confronting our National security. Federal networks have been
under attack for years. These attacks have resulted in the loss
of massive amounts of critical information, so many of these
attacks are classified.
We all know that cybersecurity is a critical National
security issue, and this committee has taken the lead. My
Ranking Member, Mr. Thompson, reintroduced his cybersecurity
bill from last year, H.R. 174, in January of this year and made
sure it was referred to this subcommittee. The need to improve
America's cyber defense posture is clear, and the Homeland
Security Committee has been arguing this point for a long time.
Now the President has come forward with a comprehensive
strategy and some legislative proposals about how it will
prevent, detect, and respond to attacks on computer systems and
infrastructure. There have been many cyber-related bills in the
last session of Congress, and the Members of Congress wrote to
the President and asked for his input on cybersecurity
legislation. As part of the President's 2-year cyberspace
policy review, the White House has put forth a detailed and
determined cybersecurity legislative proposal. I look forward
to examining that proposal today.
I thank you for calling this hearing, Mr. Chairman, and I
yield back the balance of my time.
[The statement of Ranking Member Clarke follows:]
Prepared Statement of Ranking Member Yvette D. Clarke
June 24, 2011
We live in a world where it seems that everything relies on
computers and the internet.
The effective functioning of our critical infrastructure--from
airports, financial systems, to water systems, factories, the electric
grid--is highly dependent on computer-based systems called ``control
systems'' that are used to monitor and control sensitive processes and
physical functions.
The danger of both unintentional and intentional cyber attack is
real, and the potential consequences of an attack on control systems
vary widely from the introduction of raw sewage into potable water
systems to the catastrophic failure of critical electrical generators
due to the change of a single line of code in a critical system.
We've come to recognize that public/private partnerships are a key
component of securing our Nation's computer-reliant critical
infrastructure. Private sector involvement is crucial, as it
collectively owns the vast majority of the Nation's cyber
infrastructure and is responsible for protecting its networks and
systems from the growing threat of a cyber attack.
Enhancing the public/private partnerships by developing an improved
value proposition and implementing better incentives, among other
measures, will be essential to encouraging greater private sector
involvement.
Control systems are not the only computers subject to attack. Every
day, thousands of attacks are launched against Federal and private
networks by hackers, terrorist groups, and nation-states attempting to
access classified and unclassified information, and the infiltration by
foreign nationals of Federal Government networks is one of the most
pressing issues confronting our National security.
Federal networks have been under attack for years; these attacks
have resulted in the loss of massive amounts of critical information,
though many of these attacks are classified.
We all know that cybersecurity is a critical National security
issue, and this committee has taken the lead. My Ranking Member, Mr.
Thompson re-introduced his cybersecurity bill from last year, H.R. 174,
in January of this year, and made sure it was referred to this
subcommittee. The need to improve America's cyber defense posture is
clear, and the Homeland Security Committee has been arguing this point
for a long time.
Now, the President has come forward with a comprehensive strategy,
and some legislative proposals, about how it will prevent, detect, and
respond to attacks on computer systems and infrastructure.
There have been many cyber-related bills in the last session of
Congress, and Members of Congress wrote to the President and asked for
his input on cybersecurity legislation.
As part of the President's 2-year Cyberspace Policy Review, the
White House has put forth a detailed and determined cybersecurity
legislative proposal.
I look forward to examining that proposal today, and thank you for
calling this hearing Mr. Chairman.
Mr. Lungren. I thank the gentlelady.
I now recognize the Ranking Member of the full committee,
the gentleman from Mississippi, Mr. Thompson, for any statement
he may have.
Mr. Thompson. Thank you very much, Mr. Chairman, for
holding this hearing.
I welcome our witnesses also.
Being, as you have already indicated, that there is a vote
on the way, I will submit my opening statement for the record.
[The statement of Ranking Member Thompson follows:]
Prepared Statement of Ranking Member Bennie G. Thompson
June 24, 2011
When President Obama released his Cyberspace Policy Review almost 2
years ago, he declared that the ``cyber threat is one of the most
serious economic and National security challenges we face . . .''.
I agree with him and I am pleased that his administration has taken
significant steps to put forth a clear path to update our cybersecurity
laws.
I am also pleased we are examining the President's proposal here
today.
This committee is the lead on cybersecurity in the House, as it
should be, and we have been examining this issue and calling for action
since our formation.
I re-introduced my cybersecurity bill, H.R. 174, in January of this
year with the continuing hope that it might get a hearing in this
committee.
Frankly, the White House proposal we are examining today has used
many of the concepts I suggest in my legislation.
We are facing a National and global challenge on cybersecurity, and
we must be internationally engaged to make improvements.
Simply put, we must figure out how cyberspace is to be governed,
and how it is to be secured. We know that decisions being made by
international bodies that govern the internet do not necessarily
reflect U.S. National interests.
Major corporations, financial firms, Government agencies, and
allies have all been victims of cybersecurity breaches, and these are
just the events we know about.
Classified military networks have been penetrated by foreign
intelligence agencies, and from the perpetrators' perspective, no one
has ever been punished for any of these actions. This is not a record
of success.
Since 1998, we have repeatedly tried a combination of information
sharing, market-based approaches, public/private partnership, and self-
regulation in an effort to strengthen our cyber defenses.
Hopefully, we are learning from the shortcomings of the past and
preparing for future challenges.
Mr. Chairman, I look forward to today's examination of the
President's proposal, and thank you for calling this hearing.
Mr. Lungren. I thank the gentleman for submitting his
opening statement.
We will recess until we vote and complete the vote. I
believe we just have one vote. So we will return immediately
and begin with our witnesses.
With that, this subcommittee hearing is recessed.
[Recess.]
Mr. Lungren. The subcommittee will resume.
Other Members of the subcommittee are reminded that opening
statements may be submitted for the record.
We are pleased to have a distinguished panel of witnesses
before us today on this most important topic.
Melissa Hathaway served in President Obama's
administration, 2009, where she coordinated the 60-Day
Cyberspace Policy Review. Following the report, she stood up
the Cybersecurity Office within the National security staff to
conduct work based on the blueprint. Previously, she served
under President Bush as cyber coordinator executive and
director of the Joint Interagency Cyber Task Force in the
Office of the Director of National Intelligence. Ms. Hathaway
previously worked as principal with Booz Allen & Hamilton;
currently is a strategic consultant in the field of
cybersecurity.
Mr. Greg Shannon is the chief scientist for the CERT
program at Carnegie Mellon University Software Engineering
Institute, a Federally-funded research and development center.
Mr. Shannon has previously led applied research and development
efforts in cybersecurity and data analysis at a number of
private companies as well as the Los Alamos National
Laboratory.
Leigh Williams has served as the president of BITS, the
technical policy division of The Financial Services Roundtable,
since 2007, focusing on improving operational practices and
public policy in the financial sector. Previously, he was a
senior fellow at Harvard's Kennedy School of Government,
researching public- and private-sector collaboration in the
governance of privacy and security. In addition, he has worked
at Fidelity Investments, where he was the chief risk officer
and chief privacy officer.
Then we have Mr. Larry Clinton, president and CEO of the
Internet Security Alliance, a multi-sector industry group which
was created to integrate advanced technology with the needs of
the business community, leading to a secure internet. During
his tenure at the Internet Security Alliance, Mr. Clinton
created the ``Cyber Security Social Contract.'' He has
previously worked as vice president of the USTelecom
Association and as legislative director for our former
colleague Rick Boucher, who was the subcommittee chair on the
Energy and Commerce Committee with jurisdiction over
telecommunications and the internet.
We welcome all of you. We would ask you to try and stay
within the 5 minutes. Your prepared written text will be made a
part of the record.
Before you begin, I would just ask unanimous consent that a
letter that we received from the American Chemistry Council in
regard to the subject before this committee be made a part of
the record.
Without objection, it will be.
[The information follows:]
Statement of the American Chemistry Council
June 24, 2011
acc members are a critical aspect of the nation's economy
The American Chemistry Council (ACC) represents the leading
companies in the United States who produce the chemical products
essential for everyday life. And, the business of chemistry is a
critical aspect of our Nation's economy employing more than 800,000
Americans in good-paying, high-tech positions and produces 20% of the
world's chemical products.
More than 96% of all manufactured goods are directly touched by the
business of chemistry. The chemical industry provides vital products
and materials that help improve our standard of living, advance green
energy objectives and protect the health and welfare of all Americans.
Our industry produces critical components used in lifesaving
medications, medical devices, body armor for our armed forces and law
enforcement, energy-efficient light-weight components for vehicles that
improve gas mileage, energy-saving building materials, and the durable,
light-weight wind turbine blades that help generate green energy that
creates jobs while protecting the environment.
cybersecurity is a top priority for acc and the chemical industry
Because of our critical role in the economy and our responsibility
to our communities, security continues to be a top priority for ACC
members. In 2001, our members voluntarily adopted an aggressive
security program that became the Responsible Care Security Code
(RCSC). Responsible Care implementation is mandatory for all members of
the ACC and is regularly reviewed by independent, credentialed third-
party auditors. The RCSC is a comprehensive security program that
addresses physical and cybersecurity risks. The Security Code requires
a comprehensive assessment of its cybersecurity vulnerabilities and
implementation of appropriate protective measures throughout a
company's supply chain, The RCSC has been a model for State-level
chemical security regulatory programs in New Jersey, New York, and
Maryland and was deemed equivalent to the U.S. Coast Guard's Maritime
Transportation Security Act (MTSA).
The Security Code covers the crucial area of cyber and information
security and we were gratified that in 2009 the Obama administration
made cybersecurity a top priority. Along with physical security, ACC
members began actively addressing cybersecurity issues before and after
the attacks of September 11, 2001. Cyber experts from member companies
also work closely with the DHS National Cyber Security Division (NCSD)
in many areas including: National Cyber Storm exercises, information
sharing programs, development of the Roadmap to Control Systems
Security for the Chemical Sector, A 2009 Program Update can be found on
the Obama administration's website--``Making Strides to Improve Cyber
Security in the Chemical Sector.''
Security in all its dimensions continues to be a top priority for
the ACC and the chemical industry, and we're proud of our record of
accomplishment and cooperation on cybersecurity with Congress, DHS, and
others.
the chemical industry complies with tough cybersecurity regulations
On April 9, 2007 the U.S. Department of Homeland Security published
the ``Chemical Facilities Anti-terrorism Standards'' (CFATS) regulatory
program. This comprehensive Federal regulatory program requires high-
risk chemical facilities to register with DHS, conduct a thorough site
security assessment and implement protective measures that comply with
18 risk-based performance standards (RBPS).
In particular, RBPS No. 8 establishes performance standards for
cybersecurity that must be implemented by each covered facility. RBPS
No. 8 requires facilities to deter cyber sabotage and prevent
unauthorized access to critical process control systems including
Supervisory Control and Data Acquisition (SCADA) systems, Distributed
Control Systems (DCSs), Process Control Systems (PCSs), Industrial
Control Systems (ICSs) and other sensitive computerized systems. To do
this, RBPS No. 8 requires a combination of policies and practices that
high-risk chemical facilities must address to effectively secure a
facility's cyber systems from attack or manipulation including:
(1) security policy,
(2) access control,
(3) personnel security,
(4) awareness and training,
(5) monitoring and incident response,
(6) disaster recovery and business continuity,
(7) system development and acquisition,
(8) configuration management, and
(9) audits.
In addition, CFATS specifies critical cyber systems that may
require certain enhanced security activities including those that
monitor and/or control physical processes that contain a chemical of
interest (COI); those that are connected to other systems that manage
physical processes that contain a COI; or those that contain business
or personal information that, if exploited, could result in the theft,
diversion, or sabotage of a COI.
acc recommendations for effective cybersecurity policies
ACC and its members support comprehensive cybersecurity legislation
that promotes effective collaboration between the chemical industry and
the Department of Homeland Security and ensures that robust
cybersecurity practices are implemented across the chemical supply
chain, while maintaining the free flow of commerce.
To do this ACC recommends the following:
Create cybersecurity standards that are prioritized based on
risk and focused at protecting critical systems that if
compromised would truly pose a significant threat to National
security, public safety or the National economy. Cybersecurity
legislation should establish performance standards to allow for
flexibility in their application so that chemical industry
entities can use appropriate measures that fit their unique
circumstances while ensuring the security of their critical
systems. The standards should take advantage of the incredible
wealth of knowledge embodied in the international cybersecurity
standards community.
Establish a public/private partnership to effectively share
information that is timely, specific, and actionable and is
properly protected from public disclosure. Such a partnership
will vastly improve the flow of information and ideas to help
quickly identify threats and vulnerabilities. Such an approach
will also generate flexible solutions that protect critical
cyber systems that operate complex process controls, contain
valuable intellectual property and trade secrets and personal
information on employees, customers, and suppliers. To help
promote the flow of information, information voluntarily
provided by the private sector should be adequately protected
from public disclosure including Freedom of Information Act
requests.
Provide limited liability protection for the private sector
as a result of a cyber-attack, so long as recognized
technologies have been applied to address potential threats. In
order to promote the more rapid penetration of state-of-the-art
emerging technologies to protect against cyber threats, the
Government should hold technology users harmless from damages
resulting from cyber-attacks on their IT and control systems,
so long as recognized technologies have been applied to address
potential threats. For example, the liability protections
provided by the Safety Act are appropriate to consider. This
will in turn provide the private sector better access to more
advanced and affordable end products that are safe and secure
as possible.
Strengthen U.S. laws against cybercrimes and aggressively
prosecute cyber criminals and promote international
cooperation, U.S. laws should be updated and strengthened to
protect critical infrastructure from cyber-attacks and hold
those accountable for perpetrating said acts that are intended
to cause harm to critical infrastructure operating systems,
steal intellectual property and trade secrets, or obtain
personal information for financial gain.
Consider the borderless nature of the international cyber
community and the challenges that it presents. The U.S. Federal
Government should develop strong National and international
partnerships to work together in identifying international
threats, investigate cyber-crimes, and vigorously prosecute
cyber criminals. The American chemical industry is one of the
most creative and effective manufacturing enterprises in the
world. However, with the advent of the Advanced Persistent
Threat (APT), international cyber criminals are attempting to
steal our intellectual property with little risk of getting
caught. Successful APTs could compromise our industry's ability
to compete in the global market place. Without a focused
strategy to address this issue, the private sector will
continue to fight an uphill battle. ACC encourages the Federal
Government to include this issue as a central component of its
strategy and strengthen our fight against international cyber
theft of intellectual property.
conclusion
We agree that our shared priority is to enhance cybersecurity
across the chemical supply chain Nation-wide. ACC looks forward to a
productive debate on cybersecurity legislation that protects our
critical information infrastructure while promoting effective and
efficient commerce that will continue to strengthen our economy.
The members of ACC and the chemical industry are committed to
safeguarding America's chemical facilities and the cyber systems that
enable their efficient and effective operations. It is in this spirit,
that we offer our assistance to work with the DHS and Members of
Congress in support of this shared goal.
Mr. Lungren. Ms. Hathaway.
STATEMENT OF MELISSA E. HATHAWAY, PRESIDENT, HATHAWAY GLOBAL
STRATEGIES, LLC
Ms. Hathaway. Thank you, Chairman Lungren, Ranking Member
Clarke, Members of the committee, for the opportunity to
testify on cybersecurity and its importance to homeland
security.
I am appearing today solely in my individual capacity, and
I am not representing any clients or other organizations.
Please accept my testimony for the record.
My testimony is divided into three sections: It is a review
of the threat; it is an assessment of the current legislative
docket and unaddressed needs; and a view of the need to clarify
the role for the Department of Homeland Security.
Target attacks are increasing, and our defensive posture
remains weak. Our opponents harness precision-guided bits and
bytes to deliver spam, cast phishing attacks, facilitate click
fraud, and launch a distributed denial-of-service attack. The
frequency of events and affected people and enterprises are
alarming. Recent headlines will show that our money, our
personal privacy, our infrastructure, and our children are at
risk.
The NASDAQ breach showed us that our investment plans and
money are exposed. The Epsilon breach showed us that our
personal credentials and privacy is at risk. The RSA SecureID
breach showed us that our trusted transactions and
authenticated transactions are at risk. The Sony PlayStation
network showed that our children are at risk. Then, finally,
the Stuxnet worm showed that our critical infrastructures are
at risk.
The cybersecurity problem is growing faster than the
solution. The Comprehensive National Cybersecurity Initiative,
as well as the Cyberspace Policy Review, highlighted the need
to address the threat.
Clearly, cybersecurity is a topic of interest, based on the
sheer number of bills that were highlighted in the 111th
Congress--over 55 bills--and now in the 112th Congress, showing
that a legislative conversation needs to address the shortfalls
in our current laws. As of June 2011, at least 10 pieces of
cybersecurity legislation have been introduced in the U.S.
Senate, and at least another 9 have been introduced in the U.S.
House of Representatives. I have highlighted those in my
testimony.
The cybersecurity legislative proposals reflect different
approaches and priorities. The 21st-Century digital environment
requires new laws that, at a minimum, address: Data ownership,
data handling, data protection and privacy, evidence gathering,
incident handling, monitoring and traceability, rights and
obligations related to data breach and data transfers, and
access to data based on law enforcement and intelligence
services.
The administration outlines six proposals that anchor the
priorities for debate here in Congress. As Congress considers
these proposals, it will be important to gain industry's
perspective and understand the second- and third-order effects
of these proposals.
For example, which sectors will be covered critical
infrastructure and, therefore, be subject to regulation under
the new rules? The President's international strategy for
cyberspace implies that energy, finance, transportation, and
the defense industrial base sectors will be named covered
critical infrastructures.
The proposal attempts to establish a minimum standard of
care and an audit and certification function that would be
similar in kind to the Securities and Exchange Commission
requirement for attestation of material risk. In my view,
inserting DHS into a regulatory role in this context could
dilute its operational and policy responsibilities and likely
distract from the Nation's security posture.
Additionally, the administration is proposing new
authorities for DHS by establishing a National Cybersecurity
Protection Program, which authorizes the DHS to explore
countermeasures for the overall infrastructure. The discussion
will become even more important as Congress debates the merits
of Government involvement in the protection of private-sector
networks.
As scary and as problematic as these threats are and the
intrusions may be and as devastating as they may be, it is
important that the defensive posture not overtake our core
freedoms. We should also respect the longstanding limitations
on the role of the military as it relates to public safety and
our civilian activities.
I think the most important thing that this committee can
address is whether and how we clarify DHS's overall role. Are
we going to ask them to be a policy-maker, are we going to be
asked them to be a regulator, or are we going to ask them to be
an operator?
All of the legislative proposals reflect the dilemma of a
codependent relationship between the private sector as it
develops, owns, and operates the internet-based infrastructure
for which the Government is responsible for delivering
essential services of power, water, telephone, et cetera, and
ultimately providing economic prosperity and security.
Our response includes restructuring regulation and attempts
to centralize decision-making, all with the intent to reduce
vulnerabilities and minimize the damages of intrusion. My
testimony reflects different ideas on each of the roles:
Operational, regulatory, and policy.
In conclusion, the 112th Congress has an opportunity to
drive a new legislative conversation and address the shortfalls
in current laws. The cybersecurity problem is growing faster
than the solution, and we cannot afford to be faced with
strategic surprise to address this problem. FISMA reform and a
National data breach umbrella are needed.
Additionally, modern-day criminals are using our legal
system's speed and lack thereof to their advantage. We need to
stiffen the penalties and modernize the laws that are not
keeping pace with today's digital environment. We need to
empower the National security community charged with protecting
the Nation and its critical infrastructure from cyber
exploitation or attack.
The Computer Fraud and Abuse Act, the Electronic
Communications and Privacy Act, the Stored Communications Act,
the Telecommunications Act, and the Economic Espionage Act are
among some of the laws that need to be reviewed and updated.
Congress should seek industry's perspective and debate the
advantages and challenges associated with fielding a robust and
active defense capability, imposing standards and regulation on
industry, and demanding more of DHS. An overly restrictive
approach should be avoided. We cannot afford to pass
legislation that would prove to be feckless.
I thank you very much for the opportunity to testify, and I
look forward to your questions.
[The statement of Ms. Hathaway follows:]
Prepared Statement of Melissa E. Hathaway
June 24, 2011
Mr. Chairman and Members of the committee: Thank you for the
opportunity to testify on the subject of cybersecurity and its
importance to homeland security. I am appearing today solely in my
individual capacity, and not on behalf of any clients or other
organizations.
My testimony is divided into three parts: (1) A review of the
threat, (2) an assessment of the current legislative docket and the
unaddressed needs, and (3) a view on the need to clarify the role of
DHS.
Targeted attacks are increasing and our defensive posture remains
weak.--A sense of urgency is rising because the media reports how our
insecure computers are being infected every day. Our opponents harness
precision-guided bits and bytes to deliver spam, cast phishing attacks,
facilitate click-fraud, and launch a distributed denial of service
(DDoS). The frequency of events and affected people and enterprises are
alarming. Recent headlines expose that our money, personal privacy,
infrastructure, and even our children are at risk. These network
intrusions include but are not limited to:
NASDAQ.--The operator of the Nasdaq Stock Market said it
found ``suspicious files'' on its U.S. computer servers and
determined that hackers could have affected one of its
internet-based client applications.\1\ Investigators are
considering a range of possible motives, including unlawful
financial gain, theft of trade secrets, and a National-security
threat designed to damage the exchange.\2\ Impact: Our
investment plans and money are exposed.
---------------------------------------------------------------------------
\1\ Jonathan Spicer. UPDATE 2-Hackers breach Nasdaq's computers.
Reuters On-line. 5 February 2011. http://www.reuters.com/article/2011/
02/05/nasdaq-hackers-idUSN05148621- 20110205.
\2\ Devlin Barrett. ``Hackers Penetrate Nasdaq Computers.'' The
Wall Street Journal. 5 February 2011. http://online.wsj.com/article/
SB10001424052748704709304576124502351- 634690.html.
---------------------------------------------------------------------------
Epsilon.--Epsilon, which sends 40 billion emails annually on
behalf of more than 2,500 clients, detected an incident on 30
March 2011. It determined that a subset of Epsilon clients'
customer data were exposed by an unauthorized entry into
Epsilon's email system. The information that was obtained was
limited to email addresses and/or customer names and
represented approximately 2% or 50 customers including
Walgreens, Disney destinations, Best Buy, and Citigroup.\3\ The
worry is that even months down the road, customers could get an
email impersonating their bank or credit-card issuer containing
poisonous web links. Once clicked, those links could install
malicious code on their computers or try to trick them into
giving up valuable information, such as credit card information
or log-in data to their banks or social media accounts.\4\
Impact: Our personal credentials and privacy are at risk.
---------------------------------------------------------------------------
\3\ Epsilon. Public Statement by Epsilon. 1 April 2011.
\4\ Ki Mae Heussner. Epsilon Email Breach: What You Should Know.
ABC News Online. 4 April 2011. http://abcnews.go.com/Technology/
epsilon-email-breach/story?id=13291589.
---------------------------------------------------------------------------
RSA SecureID.--In March 2011, RSA informed its customers of
a breach of its corporate network which could reduce the
effectiveness of its SecureID two factor authentication token.
On 21 May 2011, a leading U.S. defense contractor, Lockheed
Martin, had its networks penetrated. The perpetrator(s) used
duplicates of RSA's SecureID tokens to gain access to
Lockheed's internal network.\5\ After this breach and several
others resulting from the SecureID issue, RSA Security says it
will replace tokens, upon customer request.\6\ Impact: Our
trusted transactions (authenticated transactions) are at risk.
---------------------------------------------------------------------------
\5\ Jeffrey Carr. ``An Open Source Analysis Of The Lockheed Martin
Network Breach.'' Digital Dao Blog. 31 May 2011. http://
jeffreycarr.blogspot.com/2011/05/open-source-analysis-of-lockheed-
martin.html.
\6\ http://www.wired.com/threatlevel/2011/06/rsa-replaces-securid-
tokens/.
---------------------------------------------------------------------------
Sony's PlayStation Network was taken down on 20 April
2011.--A forensics team investigated the scope of the breach
and by May 2, the breach reportedly had affected an estimated
100 million people and spread to Sony's Online Entertainment
division. In an effort to show how vulnerable Sony was to a
breach, the hacker group LulzSec exposed names, birth dates,
addresses, emails, passwords, etc. of Sony's customers.\7\ As
of the end of May, Sony has spent $171 million closing the
vulnerabilities on its network and informing its customers of
their exposure.\8\ Impact: Our children are at risk.
---------------------------------------------------------------------------
\7\ Andy Bloxham. ``Sony hack: private details of million people
posted online.'' The Telegraph. 3 June 2011. http://
www.telegraph.co.uk/technology/news/8553979/Sony-hack-private-details-
of-million-people-posted-online.html
\8\ Robert Westervelt. ``Sony breach timeline shows missteps.''
Security Bytes on-line. http://itknowledgeexchange.techtarget.com/
security-bytes/sony-breach-timeline-shows-missteps-says-security-firm/
31 May 2011.
---------------------------------------------------------------------------
Citigroup.--In early June 2011, computer hackers breached
Citigroup's network and accessed the names, account numbers,
and contact data of hundreds of thousands of bankcard holders
in North America.\9\ This may be the largest breach of a
financial institution to date, arming criminals with victim
data. Impact: Our banks and money are at risk.
---------------------------------------------------------------------------
\9\ Maria Aspan. ``Regulators pressure banks after Citi data
breach.'' Reuters. 9 June 2011. http://news.yahoo.com/s/nm/20110609/
bs_nm/us_citi.
---------------------------------------------------------------------------
Stuxnet.--The Stuxnet worm that was used to shut down Iran's
nuclear program has been widely analyzed around the world. It
targets control system vulnerabilities and its source code has
been traded on the black market. Security officials worry that
this worm will be used again to attack other critical
infrastructures that rely on computers and have the same
security flaws.\10\ Impact: Our critical infrastructure is at
risk.
---------------------------------------------------------------------------
\10\ Stewart Meagher. ``Stuxnet worm hits the black market.''
THINQ. 25 November 2010. http://www.thinq.co.uk/2010/11/25/stuxnet-
worm-hits-black-market/.
---------------------------------------------------------------------------
The cybersecurity problem is growing faster than the solution.--
Upon review of these cases, it can be determined that it costs less to
break into a system or enterprise than it does to defend it. An
infected thumb drive (USB key) that costs less than $10 can undermine
an enterprise's security in minutes and nullify years' worth of
information technology (IT) investments. Organizations everywhere are
being penetrated--from small businesses to the world's largest
institutions. Policy makers, legislators, and businessmen are assessing
the gap between their current defensive posture (the floor) and their
needed front-line defense (ceiling) in the face of a growing
sophisticated range of actors. All of these facts are exasperated by
the prolonged economic recovery that has placed significant pressures
on enterprise IT budgets and focused actions toward meeting the minimum
regulatory requirements like compliance at the expense of broader
information security initiatives.
The Comprehensive National Cybersecurity Initiative (CNCI) outlined
these multidimensional threats along four attack vectors: Insider
access,\11\ proximity access;\12\ remote access;\13\ and supply chain
access \14\ and it provided a framework for unifying investments to
shore up the Government's defense. President Obama's Cyberspace Policy
Review re-stated that the Nation must become more resilient to all
types of cyber-based attacks. While there has been activity against
many of the recommendations in the Cyberspace Policy Review, there is a
lot more that needs to be done.
---------------------------------------------------------------------------
\11\ Unauthorized use or access to information, systems, and
networks by otherwise trusted agents (employees).
\12\ Gaining access to information or systems via deployment of
technology in proximity to the target.
\13\ Accessing target information and/or systems through network-
based technical means (internet).
\14\ Gaining advantage, control, and/or access to systems and the
information they contain through manipulation by cooperative/witting
vendors or unilaterally at any point in the supply chain between the
manufacturer and end-user.
---------------------------------------------------------------------------
cybersecurity in the 111th and 112th congress
The 111th Congress considered more than 50 pieces of cybersecurity
legislation. The wide range of topics addressed in these bills included
proposed changes to organizational responsibilities; instituting
compliance and accountability mechanisms; implementing data
accountability standards and reporting requirements for personal data
privacy, data breach handling and identity theft; enhancing
cybersecurity education; advancing research and development grants;
evaluating critical electric infrastructure protection and conducting
vulnerability analysis of other critical infrastructures; expanding
international cooperation on cybercrime; and addressing procurement,
acquisition, and supply-chain integrity.
Clearly, cybersecurity is a topic of interest and the sheer number
of bills highlights the cross-jurisdictional interest of the subject.
The 112th Congress has an opportunity to drive a new legislative
conversation and address the shortfalls in our current laws. As of June
2011, at least ten pieces of cybersecurity legislation have been
introduced in the United States Senate and at least another nine have
been introduced in the United States House of Representatives. Appendix
A contains a table that outlines some of the cybersecurity bills under
consideration in the 112th Congress. Like many of the bills of the
111th Congress, the bills in the 112th address niches of the
cybersecurity problems facing the Nation; even if taken together, none
of them address the situation in a comprehensive manner.
Cybersecurity legislative proposals reflect different approaches
and priorities.--The 21st Century digital environment requires new laws
that at a minimum address: data ownership; data handling; data
protection and privacy; evidence gathering; incident handling,
monitoring and traceability; rights and obligations related to data
breach and data transfers; access to data by law enforcement or
intelligence services; and degree of Government assistance (e.g.,
subsidy, information, technology, liability relief) to close the gap
between threat, innovation, and competitiveness. The Cyberspace Policy
Review identified scores of laws that needed to be updated. In May
2011, the administration put forward its cybersecurity legislative
proposal. It reflects the efforts of an interagency, consensus-based
system and a diversity of views across six proposals. Like Congress, it
shows the jurisdictional focus by specific mission areas.
Two specific areas of the administration's package have been
debated in the last two sessions of Congress: (1) Amending the Federal
Information Security Management Act (FISMA) from a static compliance-
based system to one of continuous monitoring; and (2) providing a
Federal umbrella to unify guidance of the 47 disparate State data
breach laws. The four remaining areas of the administration's package
represent new legislative proposals. Briefly, they seek to: (1) Update
the Computer Fraud and Abuse Act (CFAA) by stiffening penalties for
breaches and theft of information; (2) grant new authorities for DHS--
enabling them to deploy Intrusion Prevention Systems (IPS) in the .gov
domain and allow DHS to turn to Internet Service Providers (ISPs) to
conduct that mission on behalf of the Government (with liability
relief); (3) establish critical infrastructure regulation, set
mandatory standards for ``covered'' critical infrastructures, and an
audit and compliance regime that mandates private sector entities to
attest to cybersecurity risk management plans; and (4) prevents
restrictions on data center locations (i.e., States can't specify that
a data center be located in a certain State).
As Congress considers these proposals, it will be important to gain
industry's perspective and understand the second- and third-order
effects of the proposals. For example, which sectors will be considered
``covered'' critical infrastructure, and therefore subject to
regulation under the new rules? The President's International Strategy
for Cyberspace implies that the Energy, Transportation, Financial
Services, and Defense Industrial Base (DIB) sectors will be named the
``covered'' critical infrastructures. The legislative proposal states,
``the owners or operators of covered critical infrastructure shall
develop cybersecurity plans that identify the measures selected by the
covered critical infrastructure to address the cybersecurity risks in a
manner that complies with the regulations promulgated, and are guided
by an applicable framework designated.''\15\ This proposal attempts to
establish a minimum standard of care and an audit and certification
function that would be similar in kind to the Securities and Exchange
Commission (SEC) requirement for attestation of material risks. In my
view, inserting DHS into a regulator role in this context could dilute
its operational and policy responsibilities and likely detract from the
Nation's security posture. In May 2011, Senator Rockefeller asked the
SEC to look into corporate accountability for risk management through
the enforcement of material risk reporting.\16\ And in June 2011,
Chairman Schapiro said that the SEC would look into the matter. If
Congress believes corporations should meet such a reporting requirement
then it should turn the Executive Branch Independent Agency that is
responsible for this type of reporting and not add an additional
mission responsibility to DHS. And while regulation may be necessary,
Congress should also consider the use of other market levers (e.g., tax
relief, research and development subsidy, etc.) to incentivize industry
investment in information security.
---------------------------------------------------------------------------
\15\ The White House. Cybersecurity Legislative Package:
Cybersecurity Regulatory Framework For Covered Critical Infrastructure
Act. Page 3.
\16\ Senator Rockefeller letter to SEC Chairman Mary Schapiro. 11
May 2011.
---------------------------------------------------------------------------
Additionally, the administration is proposing new authorities for
DHS by establishing a National Cybersecurity Protection Program
(Section 244) that authorizes DHS to actively protect Federal systems.
The package states, ``the Secretary is authorized, notwithstanding any
other provision of law and consistent with section 248(a), to acquire,
intercept, retain, use, and disclose communications and other system
traffic that are transiting to or from or stored on Federal systems and
to deploy countermeasures with regard to such communications and system
traffic.''\17\ Of course more active measures must be taken to protect
Federal systems from cybersecurity threats because passive defenses are
simply not enough. The question that Congress needs to carefully
consider is which entities in the Government (e.g., Federal Bureau of
Investigation (FBI), National Security Agency (NSA), or DHS) are the
appropriate entities to help secure the Federal Government systems? Are
there appropriate checks and balances in place to oversee these new or
extended authorities?
---------------------------------------------------------------------------
\17\ The White House. Cybersecurity Legislative Package: Department
of Homeland Security Cybersecurity Authority. Page 6.
---------------------------------------------------------------------------
This discussion will become even more important as Congress debates
the merits of Government involvement in the protection of private
sector networks. The Washington Post reported last week that NSA ``is
working with internet service providers to deploy a new generation of
tools to scan e-mail and other digital traffic with the goal of
thwarting cyber-attacks against defense firms by foreign
adversaries.''\18\ Certainly other nations are turning to their ISPs as
a front line of defense in protecting their Government and private
sector networks. But, is this a mission that we want NSA to lead, or is
it one that we expect DHS to undertake?
---------------------------------------------------------------------------
\18\ Ellen Nakashima. ``NSA allies with internet carriers to thwart
cyber attacks against defense firms'' The Washington Post. 7 June 2011.
---------------------------------------------------------------------------
As scary and as problematic as these threats are and intrusions may
be (and as devastating as they may be), it is important that the
defensive posture not overtake our core freedoms. We should also
respect the long-standing limitations on the role of the military as it
relates to public safety and civilian activities. This is why, in my
opinion, the administration's legislative package proposes the section
(245) for voluntary disclosure of cybersecurity information. It
addresses shortfalls in the law and aims to extend the Provider
Exception (i.e., 18 U.S.C. 2511(2)(a)(i)) to include protection
against network attacks and prevention of delivery of malware to the
end user and provides liability relief for the reporting mechanism back
to the Government (currently not permitted under the law). One could
argue that this is what is being mandated via the code of conduct in
Australia and via the recent pan-European telecommunications reform
that will be transposed into National laws in the coming months. The
European mandate obliges the ISPs to take more responsibility for
providing enhanced security services to their customers and report all
security incidents to the European Network and Information Security
Agency (ENISA).
clarifying dhs's role: policy, operational, or regulatory
All of the legislative proposals reflect the dilemma of a co-
dependent relationship between the private sector that develops, owns,
and operates the internet-based infrastructure for which the Government
is responsible for delivering essential services (e.g., power, water,
telephone, etc.) and ultimately providing economic prosperity and
security. Our responses include organizational restructuring,
regulation, and attempts to centralize decisionmaking all with the
intent to reduce the vulnerabilities and minimize the damages of
intrusions. We appear to be asking DHS to take on new cybersecurity
roles and missions while it is establishing its basic core
competencies. Is this reasonable? Do we want DHS to become a first-
party regulator? Do we want DHS to assume an operational role that
provides actionable information to the private sector and provides
active defense of Federal systems? Or do we want DHS to assume a
broader policy role and become the National architect for a more secure
and resilient infrastructure? Perhaps it would be better to focus DHS
on becoming a center of excellence in one or two areas.
24 7 information security capability (operational)
Becoming an operational center of excellence that disseminates
timely and actionable cybersecurity threat, vulnerability, mitigation,
and warning information, including alerts, advisories, indicators,
signatures, and mitigation and response measures, to improve the
security and protection of Federal systems and critical information
infrastructure is necessary. To be successful requires DHS to adopt a
24 7 ``customer service'' business model, where its customers are other
Federal agencies; State, local, Tribal, and territorial governments;
the private sector; academia; and international partners. It would need
to learn from successful customer service industries and embed the
necessary industry partners (like the member companies of the National
Security Telecommunications Advisory Committee) within its operations.
It would need to pass knowledge onto its customers that removes the
sensitive sources and methods that make it classified and therefore
make it more readily available and actionable.
There are many other aspects of a 24 7 information security
operation that DHS could take on. Some of these capabilities are
outlined in the administration's legislative package and some
additional capabilities are outlined in other pieces of pending
legislation. Yet it is important to admit that establishing an
effective 24 7 operation is no small task. It requires real
specialization and technical expertise, a commitment to providing a
100% up-time service, and if an incident occurs, an ability to turn to
the private entities that will likely be called upon to operate in a
degraded state and restore operations (and infrastructures) quickly.
While it is possible that the National Cybersecurity and Communications
Integration Center (NCCIC) could evolve and assume this role, it would
require it to become an independent operational unit carved out of the
headquarters entity of DHS--akin to United States Secret Service or the
Drug Enforcement Agency.
If we are truly interested in setting up a 24 7 operation
immediately, then DHS in cooperation with the Department of Defense
(DoD) could call up specialist cybersecurity units within the National
Guard or DoD Reserve Forces. DHS could also turn to outside
organizations, such as the Carnegie Mellon Computer Emergency Response
Team (CERT-CC) to further augment its staff.
national architect and advocate for secure and resilient
infrastructures (policy)
Congress and the administration also turn to DHS raise awareness,
fund education initiatives, incubate technology, and broadly set
cybersecurity policies for the critical infrastructures. At the
forefront, DHS is responsible for increasing public awareness. It is
currently sponsoring a competition to develop a public service
announcement (PSA) on cybersecurity to augment the October
Cybersecurity Awareness Month. It is also conducting a review of the
university participation in the National Centers of Academic Excellence
in Information Assurance to determine how it can increase the number of
universities participating, obtain full 50-State participation,
increase the output of students per program, and align more closely
with the National Science Foundation's Scholarship for Service. Linking
these programs to hands-on experiential learning like that of the high-
school, university, and professional competitions sponsored by the U.S.
Cyber Challenge would be a natural next step.
Moreover, DHS's recently released a paper entitled, ``Enabling
Distributed Security in Cyberspace Building a Healthy and Resilient
Cyber Ecosystem with Automated Collective Action'' that explores the
idea of a healthy, resilient--and fundamentally more secure--cyber
ecosystem of the future. It envisions an environment of cyber
participants, including cyber devices, that are able to work together
in near-real time to anticipate and prevent cyber attacks, limit the
spread of attacks across participating devices, minimize the
consequences of attacks, and recover to a trusted state.\19\ If DHS
were to drive the implementation of this vision it will require DHS to
modify its relationship with industry, consolidate the number of
private-public partnerships, and drive the development of standards in
partnership with the National Institute of Standards and Technology
(NIST). It will also require DHS to lead the discussion on behalf of
the Executive Branch for the following questions: ``What are the
business drivers that will incentivize the necessary investments? What
are the appropriate roles and responsibilities of the public and
private sector in delivering the healthy ecosystem? Which elements
should be prioritized for early realization? As a healthy cyber
ecosystem emerges, governance questions become salient. Will system
owners cede decisionmaking to the community? Who sets policy for inter-
enterprise information exchange and deployment of countermeasures? What
liability regimes apply for collateral consequences of countermeasure
deployment (or the failure to deploy known countermeasures)? What legal
authorities should local and National governments, as well as
international entities, have to compel action by devices owned by or
serving private parties in order to secure the larger cyber
commons?''\20\
---------------------------------------------------------------------------
\19\ Department of Homeland Security. ``Enabling Distributed
Security in Cyberspace Building a Healthy and Resilient Cyber Ecosystem
with Automated Collective Action.'' 23 March 2011.
\20\ Department of Homeland Security. ``Enabling Distributed
Security in Cyberspace Building a Healthy and Resilient Cyber Ecosystem
with Automated Collective Action.'' 23 March 2011. Page 27.
---------------------------------------------------------------------------
Like the operational role, this policy-based role requires
personnel who are steeped with background in policy development and the
art of negotiation. It also requires understanding of the technical
underpinnings of the next generation hardware and software and
knowledge of the standards-setting processes. Raising awareness and
advocating a new architecture of hardware and software products for
industry to build toward is no small task. If Congress and the
administration want DHS to be the National voice for cybersecurity,
they cannot necessarily be saddled with all of the operational and
regulatory missions that are recommended in the legislative proposals.
first-party regulatory role vice-setting standards
Is it possible for regulation to keep pace with technology
development and adoption? Has the market failed to produce secure and
resilient hardware and software products?
Many of the critical infrastructures are already regulated (e.g.,
energy, finance, telecommunications) and NIST works with the Sector
Agency and DHS to set the standards by which industry has to meet. But
as evidenced by the three volume edition on Guidelines for Smart Grid
Cybersecurity,\21\ the standards are not always published in time for
market penetration and adoption. So, what is the role of the private
sector in policing itself, adapting to new industry standards and
upgrades, and coping with accelerating threats? The North America
Electric Reliability Corporation (NERC) works across the electric power
sector to set the standards and help ensure compliance. However, due to
the intermingling of State and Federal regulation the industry usually
adopts a lower standard leaving some vulnerabilities unaddressed.
Existing standards will never be sufficient in light of a
sophisticated, perhaps nation-state adversary, but they can be
strengthened.
---------------------------------------------------------------------------
\21\ Department of Commerce, National Institute of Standards and
Technology. Guidelines for Smart Grid Cybersecurity (3 volumes). August
2010.
---------------------------------------------------------------------------
What may be more useful would be if DHS, supported by the FBI and
intelligence community, were to inform industry of the threats they are
facing and how they are being exploited or penetrated. A training
program that educates corporate leadership on how to mitigate the risk
of being a high-value target including providing them with briefings
about the threat to their industry using specific case studies may go
along way to reducing the number of incidents and loss of confidential
information. Furthermore, as some companies are ``personally'' touched
by the penetration of their networks (e.g., Sony and Citigroup), they
may be extra motivated to invest in and promote stronger information
security standards for their industry and customers alike.
As Congress considers placing DHS into more of a regulatory role,
it should consider the impact of the possible dilution of its
operational and policy responsibilities. While some say DHS's input and
support of streamlining CIP standards has had a positive affect, is it
making enough of a difference? Is it best to educate the first-party
regulators and help them improve the security posture of the Nation?
How are the other existing regulatory bodies (SEC, FCC, FERC, or FTC)
using their current authorities to address the situation? Would
strengthening the regulatory oversight of the SEC, FCC, FERC, or FTC
help or hurt the situation?
conclusion
The 112th Congress has an opportunity to drive a new legislative
conversation and address the shortfalls in our current laws. The
cybersecurity problem is growing faster than the solution and we cannot
afford to be faced with strategic surprise to address the problem.
FISMA reform and a National data breach umbrella are needed.
Additionally, modern-day criminals are using our legal systems' speed,
or lack thereof, to their advantage. We need to stiffen penalties and
modernize the laws that are not keeping pace with today's digital
environment. We need to empower the National security community charged
with protecting the Nation and its critical infrastructure from cyber
exploitation or attack. The Computer Fraud and Abuse Act, Electronic
Communications and Privacy Act, Stored Communications Act,
Telecommunications Act, and Economic Espionage Act are among some of
the laws that need to be reviewed and updated. Congress should seek
industry's perspective and debate the advantages and challenges
associated with fielding a robust active defense capability, imposing
standards and regulation on industry, and demanding more of DHS. An
overly restrictive approach should be avoided yet, we cannot afford to
pass legislation that would prove to be feckless.
I thank you very much for the opportunity to testify, and look
forward to your questions.
Exhibit A
REVIEW OF CYBERSECURITY LEGISLATION IN THE 112TH CONGRESS
------------------------------------------------------------------------
United States House of
United States Senate Representatives
------------------------------------------------------------------------
S. 8, Tough and Smart National Security H.R. 76, Cybersecurity
Act. Education Enhancement Act of
2011.
S. 21, Cyber Security and American H.R. 96, Internet Freedom Act
Cyber Competitiveness Act of 2011. of 2011.
S. 28, Public Safety Spectrum and H.R. 174, Homeland Security
Wireless Innovation Act. Cyber and Physical
Infrastructure Protection Act
of 2011.
S. 372, Cybersecurity and Internet H.R. 607, Broadband for First
Safety Standards Act. Responders Act of 2011.
S. 413, The Cybersecurity and Internet H.R. 668, Secure High-voltage
Freedom Act of 2011. Infrastructure for Electricity
from Lethal Damage Act (SHIELD
Act).
S. 709, Secure Chemical Facilities Act. H.R. 1136, Executive Cyberspace
Coordination Act of 2011.
S. 813, Cyber Security Public Awareness H.R. 1389, Global Online
Act of 2011. Freedom Act of 2011.
S. 968, Preventing Real Online Threats H.R. 1540, National Defense
to Economic Creativity and Theft of Authorization Act for Fiscal
Intellectual Property Act of 2011 Year 2012.
(PROTECT IP Act).
S. 1101, Electronic Communications and ...............................
Privacy Act--Amendments Act (Digital
Privacy Bill).
S. 1151, Personal Data Privacy and ...............................
Security Act of 2011.
------------------------------------------------------------------------
Mr. Lungren. Thank you very much for your testimony.
Now Dr. Shannon.
STATEMENT OF GREGORY E. SHANNON, CHIEF SCIENTIST FOR COMPUTER
EMERGENCY READINESS TEAM (CERT), SOFTWARE ENGINEERING
INSTITUTE, CARNEGIE MELLON UNIVERSITY
Mr. Shannon. Thank you, Chairman Lungren, Ranking Member
Clarke, and other Members of the subcommittee, for me to talk
about, this morning, the future of cyber incident response. I
applaud the current efforts of Congress to mitigate risks to
our public and private critical information infrastructures.
CERT, as you mentioned, is a Federally-funded Department of
Defense research and development lab. We have over 250 staff
that have been working on this challenge of incident response
since 1988, when the Morris worm first was experienced. For
example, we catalogue over a quarter-million malware artifacts
each month. We assist in major, on-going cybersecurity
incidents of National importance. We release security coding
guidelines and technologies for the C, C++, and Java
programming languages.
While much is said about risk mitigation, incident response
receives less focused attention as a strategic technical area,
yet it is critically important. Vigorous attacks on our network
environments will continue for the foreseeable future, failures
will occur, and effective responses are required. The Federal
Government must look at incident response as strategic, just as
it looks at preventative efforts. The U.S. CERT and other
capabilities are a part of this effort.
Our country needs legislation that will facilitate capable,
scalable, cost-effective cybersecurity incident response for
critical Government infrastructure. Things will fail in
unexpected ways, and our Nation must have the capacity to
respond accordingly.
I believe that the most difficult technical challenge to
both effective risk mitigation and incident response is
selecting practices that are scientifically sound and
operationally proven. We do not want to be guessing. I
encourage you to consider in the rulemaking language that valid
approaches be considered. The complexity of practices and
regimes being proposed will probably have unintended and
unexpected consequences. Some approaches aren't fully proven,
experimentally or operationally. Again, I encourage you to use
language that calls that out in the rulemaking.
I believe that the most difficult policy challenge for
effective Government incident response is harmonizing the
responsibilities, authorities, capabilities, and communication
across the various agencies, as Ms. Hathaway has highlighted.
At CERT, part of our value to the Government has been the
ability to bridge these gaps and misalignments in the midst of
the response to a critical cybersecurity incident. But we
recognize that that is not the ideal way, going forward, to be
ad hoc.
Three areas that we highlight in the written testimony is:
Data sharing, forensics, and training. I would encourage--I
applaud the effort of safe harbors in Section 246 for
organizations and individuals that are attempting to do the
right thing. The notion of ``right thing'' in incident response
is a well-founded principle that individuals, organizations
often know what the right thing to do is, and it is important
that the policies and such be aligned to support that.
On the forensics side, what we are seeing is an excellent
use of potential cloud-based computing, private clouds, to
support a broad capability for the law enforcement community to
do investigations at scale. As these incidents increase in
scope and scale, the ability to respond quickly with
appropriate forensics, to maintain the velocity of the
investigation, as well as to collect the evidence that could be
used in court, is important.
Finally, on training, one of the key challenges is how to
train as the technical people work, or, as the Department of
Defense says, train as you fight. The environments that we are
in are complex. The threats that are experienced are even more
complex and less likely to be experienced. Part of the work we
do is to encourage the ``train as you work'' mentality, to be
realistic.
We at CERT look forward to the day when our Nation's
cybersecurity resiliency is founded on the effective mitigation
of cyber risks and pervasive capabilities to respond to
cybersecurity incidents. I see this legislation and the related
modifications and efforts as an important step in the right
direction.
For your benefit, I would like to also submit an article
from Nature. It talks about the Stuxnet. This was in the June
issue. At the end of it, it highlights some of the technical
challenges from a science-of-security point of view. I would
like to also submit that into the written record.
Mr. Lungren. Without objection, that shall be accepted.*
---------------------------------------------------------------------------
* The information has been retained in committee files and is
available at http://www.nature.com/news/2011/110608/full/474142a.html.
---------------------------------------------------------------------------
Mr. Shannon. Okay. Thank you for your time.
[The statement of Mr. Shannon follows:]
Prepared Statement of Gregory E. Shannon
June 24, 2011
Chairman Lungren, Ranking Member Clarke, and other distinguished
Members of the subcommittee, thank you for the opportunity to testify,
it is my pleasure to be here this morning to discuss cyber incident
response.
about cert
The CERT Program is part of Carnegie Mellon University's Software
Engineering Institute, a Federally-funded research and development
center, and is located on the Carnegie Mellon campus in Pittsburgh,
Pennsylvania.
The CERT program (http://www.cert.org/) was charged by DARPA in
1988 to set up the first Computer Emergency Response Team (CERT) as a
response to the Morris worm incident. We continue to develop and
promote the use of appropriate technology and systems management
practices to resist attacks on networked systems, limit damage, and
restore continuity of critical services. CERT works both to mitigate
cyber risks and coordinate cyber incident responses at local, National,
and global levels. Over the last 23 years CERT has helped to establish
over 200 CERT computer security incident response teams (CSIRTs) around
the world--including the DHS US-CERT. We continue to have proven
success transitioning research and technology to those who can
implement it on a National scale.
Dr. Greg Shannon is the Chief Scientist for the CERT Program, where
he works to establish and enhance the program's research visibility,
initiatives, strategies, and policies.
testimony
Today's operational cyber environments are complex and dynamic.
User needs and environmental factors are constantly changing, which
leads to unanticipated usage, reconfiguration, and continuous evolution
of practices and technologies. New defects and vulnerabilities in these
environments are continually being discovered, and the means to exploit
these environments continues to rise. The CERT Coordination Center
cataloged 250,000 instances of malicious artifacts last month alone.
From this milieu, public and private institutions respond daily to
repeated attacks and also to the more serious previously un-experienced
failures (but not necessarily unexpected); both demand rapid, capable,
and agile responses.
Incident response, as a discipline, is maturing. Over the last two
decades, it has emerged from the shadows of IT and risk management, to
achieve recognition as a robust and growing discipline.\1\ Signs of
this progress include the emergence of process models, meta-models,
bodies of knowledge, common data representations, and auditable
standards. Further development, and continued funding, will enable
faster and more efficient dissemination of information to trusted
partners in larger trust networks.
---------------------------------------------------------------------------
\1\ For example, this fall, CERT and the Institute for Information
Infrastructure will hold a workshop on Coordinated Private-Sector
Responses to Cyber Security Incidents. This is a follow on to I3P's
2009 workshop on Protecting Critical Infrastructures: The National
Capital Region as a Model for Cyber Preparedness.
---------------------------------------------------------------------------
I applaud the current efforts of the Federal Government to mitigate
risk to our public and private critical information infrastructures;
CERT has worked tirelessly to improve cybersecurity in areas such as
secure coding, insider threat, and vulnerability analysis. But, while
much is said about risk mitigation, incident response is often not as
thoroughly addressed, and is critically important. Networked
environments will continue to be vigorously attacked for the
foreseeable future. Failure will occur and effective responses are
required. Incident response is not a single action but rather a complex
function that includes containment, repair, and recovery.\2\ The
Federal Government must look at incident response as strategic, just as
it looks at preventative efforts. Our country needs legislation that
will facilitate capable, scalable, and cost-effective cyber-incident
response for critical and Government infrastructure. Things will fail
in unexpected ways and our Nation must have the capacity to respond
accordingly.
---------------------------------------------------------------------------
\2\ Some contend that retaliation is part of incident response; I
disagree. The response community does not consider it in scope for
incident response as practiced today. Other organizations and
disciplines are better suited to address this issue.
---------------------------------------------------------------------------
I believe that the most difficult technical challenge to effective
risk mitigation and incident response is selecting practices that are
scientifically sound and operationally proven. The complexity of
practices and regimes being proposed in legislation and elsewhere will
probably have unintended and unexpected consequences. I encourage the
subcommittee to use language in legislation that encourages practices
that are both experimentally and operationally validated.
I believe that the most difficult policy challenge for effective
Government incident response is harmonizing the responsibilities,
authorities, capabilities, and communication across the various
agencies involved. I support the current efforts in this.
In my remaining testimony I discuss three areas that we at CERT
believe are key to the future of incident response.
information sharing
We all realize how critical it is for stakeholders to share
information, but good incident response is contingent upon sharing the
right information, with the right people, at the right time. High-
quality and actionable information comes from superior situational
awareness only possible with robust information sharing and sufficient
visibility into one's own enterprise. Currently, our technical
capabilities allow us to see and respond to variant indicators, but to
better detect, share, and respond to incidents analysts need to be able
to look past narrowly-focused indicators.
Achieving this enhanced situational awareness will require
continued research on network traffic and data. The ability to detect
malicious markers that are invariant, such as behavioral-based
indicators (e.g. insider threats) will enable a more proactive
response. To facilitate innovation, richer data needs to be shared with
the research community, not only incident data itself, but also data-
sets that will enable an understanding of what ``normal'' resembles.
Currently, the community does not have a clear understanding of what
this data set would look like. If situational awareness is to develop
beyond simple indicators, regulatory frameworks must allow access to
everyday data, so that investigators can begin to recognize what data-
set are important. This data sharing should start with limited access
to high-fidelity data sets for researchers so that data with
scientifically proven value is considered for sharing operationally.
Otherwise, policymakers and experts are left to speculate what is the
right data to share. To further improve the future efficiency and
effectiveness of incident response, the community also needs to develop
and use automated tools and techniques to analyze and correlate the
vast amount of log files, artifacts, and other event information.
Moreover, compliance-driven information sharing will only lead to
the bare minimum disclosure of sensitive information related to
problems, concerns, and vulnerabilities. Building trusted relationships
with stakeholders becomes essential to avoiding such limited
information exchange and is a fundamental ingredient to a successful
response. We also have to trust the people in the field and those who
first respond to incidents. I applaud the effort in this legislation to
support actions to do the ``Right ThingTM''; this is an
important principle in the response community and is the basis of
successful responses in many highly stressful incidents. Safe harbor
measures such as Sec. 246 in the administration's Cybersecurity
Legislative Proposal work towards continued encouragement to share
data; however in response scenarios it is worthwhile to consider
including the actions of cyber ``first responders'' into good faith
legislation as well.
forensics
While gains have been made in the field of incident response the
nature of the ever-evolving cyber threat poses a huge challenge and
demand for incident response expertise that has far outstripped the
supply.
Computers are no longer just the targets of crime; our adversaries
now use them to facilitate every aspect of their illicit activities and
achieve effects at scale. Once an incident occurs Federal agencies are
facing several hurdles to recover the needed data in order to locate
the source of the incident and contain the problem. First, computer
forensic labs are constrained by a lack of resources, creating an
enormous backlog rendering them unable to handle the megafold increases
in the volumes of data that need to be examined for evidence. While
some agencies may have the qualified examiners, and many do not, they
lack the funds to properly equip them for the mission. For example,
current examination methods rely heavily on processor power, but due to
dramatically increased computer memory, examination stations often
cannot keep up. Finally, the current state of the practice does not
allow examiners to easily access varied levels of expertise in a timely
or cost-effective way; frequently people are sent Temporary Duty or
images are shipped to higher level units, resulting in time delays and
increased costs.
To successfully respond to cyber incidents these obstacles must be
overcome in a way that allows for high-quality collaborative
examinations. For instance, what would happen if an adversary
perpetrated an actual, severe cyber event with National consequences?
Currently there is no one facility or lab that could support the volume
of data these kinds of events would generate. Under current conditions,
data would have to be distributed, adding to the time and complexity of
conducting examinations. Agencies will need to augment scarce resources
by having multiple users viewing the same data either remotely or
locally, while maximizing the application of specialized computing
resources, and allowing for massive, coordinated efforts. Analysts and
investigators will need flexible, secure access to high-performance
systems, to increase productivity and facilitate effective distributed
collaboration in a scalable and cost-effective way.
training
In order to rapidly handle cyber incidents the Federal Government
needs a workforce educated and equipped to respond. However, the rapid
changes and dynamic nature of cybersecurity make keeping the workforce
up to date a very challenging problem. Responding to critical cyber
events requires technical knowledge and skills, decision-making
abilities, and effective coordination--all while moving rapidly.
Moreover, a lack of preparation inhibits secondary incident-handling
activities, such as: Evidence gathering, identifying the attacker, and
reporting the incident to other affected organizations. The Federal
Government must have an agile and prepared workforce to deal with cyber
incidents, and should to be able to train them in a cost-effective and
scalable manner.
The most common workforce development training solution is the
traditional classroom training model. While this training model is easy
to implement and is widely used, there are a number of reasons why it
is not adequate for providing effective, large-scale training to a
technical workforce, including time, cost, and scalability.
Furthermore, traditional classroom training is not optimal for rapidly-
changing fields such as cybersecurity.
The best way to prepare the workforce is to have them practice
under realistic conditions with interactive simulations, and the
ability to interface with participants across multiple locations who
can work together to analyze and respond to the latest threats and
attacks. Individuals need to be trained on a platform that safely
mimics how the internet would respond to stress and exposes them to
real-world scenarios, events, and activities that are similar to those
they will encounter in their jobs.
In addition, there are two incident response domains where we see
an immediate need for further training. The first is reverse
engineering, to grow capacity in analyzing malware recovered from an
incident. The second domain is embedded systems, which pose many unique
challenges for incident response and which some experts believe will be
a major cybersecurity problem area in the near future.
The workforce needs to not only be trained, but also educated. For
example, in the case of forensics, much of the training the workforce
receives is how to use tools, but when those tools are not effective no
one is educated on how to manage the situation or apply critical
thinking to determine alternative approaches. What's more, to train the
workforce to manage cyber incidents the Federal Government needs to
expand the scope of computer or cybersecurity training to include first
responder training and best practice guidance. Without proper education
a first responder may unintentionally cause irrevocable damage by doing
something as simple as turning off a computer. This will not only cause
lost data, but can also result in severely slowing an investigation and
compromise the potential prosecution of the perpetrator.
In conclusion, I thank the subcommittee again for inviting me and
considering my testimony. Our Nation will continue to see significant
serious cyber incidents for the foreseeable future. CERT's mission is
to help ensure that these incidents are not catastrophic and that we
recover as quickly as possible. We at CERT look forward to the day when
our Nation's cyber resiliency is founded on the effective mitigation of
cybersecurity risks and pervasive capabilities to respond to
cybersecurity incidents. I see this legislation and the related
modifications and efforts as an important step in the right direction.
Mr. Lungren. Mr. Williams.
STATEMENT OF LEIGH WILLIAMS, PRESIDENT, BITS, THE FINANCIAL
SERVICES ROUNDTABLE
Mr. Williams. Thank you, Chairman Lungren, Ranking Member
Clarke, and Members of the committee.
My name is Leigh Williams, and I am president of BITS, the
technology policy division of The Financial Services
Roundtable, where we address security fraud and other
technology issues on behalf of our 100 member institutions,
their millions of customers, and all of the stakeholders in the
U.S. financial system.
In my remarks today, I will briefly describe cybersecurity
in financial services, explain why The Roundtable supports the
Obama administration's cybersecurity legislation, and comment
on some of the strong provisions of H.R. 174.
In my view, most cybersecurity protection arises from
individual institutions investing literally tens of billions of
dollars and tens of millions of hours in voluntary measures for
business reasons. Up at the industry level, BITS and several
other coalitions promote best practices for protecting customer
information. For example, BITS is currently addressing security
in mobile, cloud computing, social networking, protection from
malicious software, and security training and awareness.
Beyond these voluntary efforts, our members are also
subject to a range of oversight mechanisms to ensure
consistency throughout the industry. Just to take security and
privacy provisions of Gramm-Leach-Bliley as an example,
Congress enacted GLB; the banking regulators detailed it in Reg
P; Reg P was translated into examination guidance; banks used
that guidance to manage their risk and the risk of their
service providers; examiners audit the banks against it;
Treasury monitors their consistency; and then just to bring
this whole process full circle, the Congress oversees Treasury
and the agencies.
Beyond this sector-specific work, we collaborate more and
more with DHS, with law enforcement, with the intelligence
community, and with other industries on a variety of projects,
including one that we have launched recently with DHS, the
Cyber Operational Resiliency Review, where institutions can
invite DHS to review their control practices and their network
traffic.
As the committee considers action on cybersecurity, I would
urge Members to appreciate these current safeguards and these
existing collaborations so that we might leverage all of them
for maximum benefit.
Even given this headstart, we believe that comprehensive
cybersecurity legislation is warranted. It can improve security
throughout the cyber ecosystem, including in telecom networks,
in software and hardware supply chains, in Federal systems, and
in our sector.
Specifically, The Roundtable supports the administration's
legislative proposal. We support many of the provisions on
their individual merits, and we see the overall proposal as an
important first step in building a more integrated approach.
We do believe that harmonizing the comprehensive approach
and the sector-specific mechanisms will be a challenge. There
are at least a couple of ways of bridging this ecosystem sector
divide. First, Congress could establish uniform standard but
with exceptions where substantially similar requirements
already are in place, as in the banking regulators' breach
notification rules. Or Congress could reserve more autonomy for
the sectors. For example, it could be the sector-specific
agencies, and not DHS, that designate the critical sector
entities or systems or assets.
In other specific provisions of the proposal, we support
strengthening penalties for computer crime, including the theft
of intellectual property. We support a uniform national
standard for breach notification with strong preemption. And we
support the Federal systems provisions, both to safeguard the
data that we report and to the systems and because we believe,
as the Chairman has suggested, that Government should use its
procurement power to model good behavior.
On H.R. 174, the Homeland Security Cyber and Physical
Infrastructure Protection Act, we see two more promising
options for harmonizing DHS and sector-level work. DHS can
delegate authority to the sector, and DHS is instructed to use
the primary regulators as conduits to the covered companies.
With these options, delegation and conduit, and the options in
the administration proposal already in place, and sector plus
aggregation, we should be able to take full advantage of both
the sector and DHS. Finally, we appreciate H.R. 174's focus on
risk-based performance-based regulation, on R&D, and on
information-sharing among the critical companies and key
agencies.
In conclusion, may I just say that at The Financial
Services Roundtable we will continue to strengthen security
around our customers' information, we will help answer the
question of ecosystem sector balance, and we will support and
we will work to implement the administration's cybersecurity
proposal.
Thank you very much for your time.
[The statement of Mr. Williams follows:]
Prepared Statement of Leigh Williams
June 24, 2011
Thank you Chairman Lungren, Ranking Member Clarke, and Members of
the committee for the opportunity to testify before you today.
My name is Leigh Williams and I am president of BITS, the
technology policy division of The Financial Services Roundtable. BITS
addresses issues at the intersection of financial services, technology,
and public policy, on behalf of its 100 member institutions, their
millions of customers, and all of the stakeholders in the U.S.
financial system.
From this perspective, I will briefly describe cybersecurity and
data protection in financial services, including private sector
efforts, sector-specific oversight and inter-sector interdependencies.
I will explain why The Financial Services Roundtable supports the
cybersecurity proposal delivered by the Obama administration to the
Congress on May 12. Finally, I will comment on the key provisions of
H.R. 174, which I understand is under active consideration by the
committee.
financial institutions' voluntary cybersecurity efforts
Within the financial services sector, the greatest amount of
cybersecurity protection arises from voluntary measures taken by
individual institutions for business reasons. To protect their retail
customers, commercial clients and their own franchises, industry
professionals--from Chief Information Security Officers to CIOs to
CEOs--are increasingly focused on safeguards, investing tens of
billions of dollars in data protection. They recognize the criticality
of confidentiality, reliability, and confidence to their success in the
marketplace. This market-based discipline is enforced through an
increasingly informed consumer base, and by a very active commercial
clientele that often specifies security standards and negotiates for
audit and notification rights.
At the industry level, BITS and several other coalitions facilitate
a continuous process of sharing expertise, identifying and promoting
best practices, and making these best practices better, to keep pace in
a dynamic environment. For example, as BITS and our members implement
our 2011 business plan, we are addressing the following items
associated with protecting customer data:
Security standards in mobile financial services.
Protection from malicious or vulnerable software.
Security in social media.
Cloud computing risks and controls.
Email security and authentication.
Prevention of retail and commercial account takeovers.
Security training and awareness.
While much of this institution-level and industry-level effort is
voluntary--not driven primarily by regulation--it is not seen by
industry executives as discretionary or optional. The market, good
business practices and prudence all require it.
oversight
To strengthen public confidence and to ensure consistency across a
wide variety of institutions, Federal financial regulators codify and
enforce an extensive system of requirements. Many of these represent
the distillation of previously voluntary best practices into
legislation introduced in Congress, enacted into law, detailed in
regulation, enforced in the field, with feedback to the Congress in its
oversight capacity.
In addition to these Federal authorities, institutions are subject
to self-regulatory organizations like the Financial Industry Regulatory
Authority (FINRA), State regulators like the banking and insurance
commissioners, independent auditors, outside Directors, and others.
These various oversight bodies, for example, apply the Financial
Services Modernization Act of 1999 (GLB), the Fair and Accurate Credit
Transactions Act (FACTA), Electronic Funds Transfers (Regulation E),
Suspicious Activity Reporting (SARs), the International Organization
for Standardization criteria (ISO), the Payment Card Industry Data
Security Standard (PCI), BITS' own Shared Assessments and many, many
more regulations, rules, guidelines, and standards.
inter-sector collaboration
Commensurate with the escalating cybersecurity challenges and
increasing interconnectedness among sectors, more and more of our work
entails public/private and financial/non-financial partnerships. Our
Financial Services Sector Coordinating Council (FSSCC) of 52
institutions, utilities, and associations actively partners with the 17
agencies of the Finance and Banking Information Infrastructure
Committee (FBIIC). [For additional detail on the FSSCC's perspective on
cybersecurity, research and development, and international issues,
please refer to the April 15, 2011 testimony of FSSCC Chair Jane Carlin
before this subcommittee.] Our Financial Services Information Sharing
and Analysis Center (FS-ISAC) is in constant communication with the
Department of Homeland Security (DHS), law enforcement, the
intelligence community and ISACs from the other critical infrastructure
sectors, to address individual incidents and to coordinate broader
efforts.
Other examples of collaboration with non-financial partners, drawn
just from BITS' 2011 agenda, include:
The Cyber Operational Resiliency Review (CORR) pilot, in which
institutions may voluntarily request Federal reviews of their
systems, in advance of any known compromise--with DHS and the
Treasury.
Multiple strategies for enhancing the security of financial
internet domains--with the Internet Corporation for Assigned
Names and Numbers (ICANN) and Verisign, in partnership with the
American Bankers Association (ABA) and in consultation with
members of the Federal Financial Institutions Examination
Council (FFIEC).
A credential verification pilot--with DHS and the Department of
Commerce--building on private sector work that began in 2009,
was formalized in a FSSCC memorandum of understanding in 2010,
and was featured in the April 15, 2011 announcement of the
National Strategy for Trusted Identities in Cyberspace (NSTIC).
Through the processes and initiatives above and in many other
efforts, financial institutions, utilities, associations, service
providers and regulators continue to demonstrate a serious, collective
commitment to strengthening the security and resiliency of the overall
financial infrastructure. As the committee considers action on
cybersecurity, I urge Members to be conscious of the protections and
supervisory structures already in place and the collaborations
currently underway, and to leverage them for maximum benefit.
need for legislation
Even given this headstart and substantial momentum, we believe that
cybersecurity legislation is warranted. Strong legislation can catalyze
systemic progress in ways that are well beyond the capacity of
individual companies, coalitions, or even entire industries. For
example, comprehensive legislation can:
Raise the quality and consistency of security throughout the full
cyber ecosystem, including the telecommunications networks on
which financial institutions depend.
Enhance confidence among U.S. citizens and throughout the global
community.
Strengthen the security of Federal systems.
Mobilize law enforcement and other Federal resources.
Enable and incent voluntary action through safe harbors and
outcome-based metrics, rather than relying primarily on static
prescriptions.
Attached are a list of 13 policy approaches that the FSSCC recently
endorsed, along with three that it deemed problematic. We urge the
committee to consider the FSSCC's input, particularly in light of the
FSSCC's leadership of the financial services industry on this issue.
administration proposal
On May 12, 2011, on behalf of the administration, the Office of
Management and Budget transmitted to Congress a comprehensive
legislative proposal to improve cybersecurity. The Financial Services
Roundtable supports this proposal and looks forward to working for its
passage. We support many of the provisions of this proposal on their
individual merits, and we see the overall proposal as an important step
toward building a more integrated approach to cybersecurity. Given that
our member institutions operate Nationally, are highly interdependent
with other industries, and are already closely supervised by multiple
regulators, we appreciate that this proposal promotes uniform National
standards, throughout the cyber ecosystem, with the active engagement
of sector-specific agencies and sector regulators.
Consistent with its comprehensive approach, the proposal strives to
address cybersecurity both at the level of the entire ecosystem and
also within specific sectors. For example:
The DHS Cybersecurity Authority title naturally stresses DHS' role,
but it also mentions ``other relevant agencies'' and sector
coordinating councils.
The Regulatory Framework title focuses largely on DHS leadership
and standardized evaluations, but it also mentions ISACs and
sector-specific regulatory agencies, and provides for sector-
level exemptions.
We believe that harmonizing the comprehensive approach with the
need to incorporate sector-specific mechanisms will be one of the most
important challenges as the Congress considers this proposal. As this
committee considers DHS' role, and its relationship to the sector-
specific roles, we urge Members to leverage existing financial services
protections and circumstances, and their analogs in other sectors,
while preserving the inter-sector quality of the proposal. Below, we
offer the committee two potential approaches and illustrations for
addressing this DHS/sector nexus:
Establish a uniform standard with specified exceptions.--In
the Data Breach Notification title, the Federal Trade
Commission (FTC) could enforce the requirements enacted under
this bill, but defer to sector-specific regulators where
substantially similar sector-specific rules and guidelines
already are in place (e.g. the FFIEC could continue to enforce
its 2005 interagency breach response guidance, and the
Department of Health and Human Services could continue to
enforce HITECH).
Preserve sector autonomy with centralized information
aggregation and coordination.--In the Regulatory Framework
title, rather than requiring DHS to list critical
infrastructure entities for every sector, the sector-specific
agencies could make that determination, just as the Financial
Stability Oversight Council is responsible for designating
Systemically Important Financial Institutions.
Given the likely fluidity of the overall solution, we cannot yet
make a definitive recommendation for either approach. We do believe
that this question of ecosystem/sector balance warrants careful
deliberation.
Law Enforcement
We support the proposal's clarification and strengthening of
criminal penalties for damage to critical infrastructure computers, for
committing computer fraud, and for the unauthorized trafficking in
passwords and other means of access. We also urge similar treatment for
any theft of proprietary business information. With this extension to
intellectual property, the law enforcement provisions will improve
protections for both consumers and institutions, particularly when
paired with expanded law enforcement budgets and the recruitment of
personnel authorized in later titles. For purposes of this title and
others, we presume that many, but not all, financial services systems
and entities will be designated as critical infrastructure vital to
National economic security, and we look forward to further work on the
associated criteria.
Data Breach Notification
We support the migration to a uniform National standard for breach
notification. Given existing State and financial services breach
notification requirements, this migration will require both strong pre-
emption and reconciliation to existing regulations and definitions of
covered data. [Please see the 2005 FFIEC Interagency Guidance on
Response Programs for Unauthorized Access to Customer Information and
Customer Notice.] We support the exemptions for data rendered
unreadable, in breaches in which there is no reasonable risk of harm,
and in situations in which financial fraud preventions are in place.
DHS Authority
We support strengthening cybersecurity authorities within DHS--and
the active collaboration of DHS with the National Institute of
Standards and Technology (NIST), sector-specific agencies such as the
Treasury Department, and sector regulators such as our banking,
securities, and insurance supervisors. This title demonstrates both the
administration's commitment to an integrated approach and the challenge
of achieving it. Federal and commercial systems, financial and non-
financial information, DHS planning and sector coordinating council
collaboration, are all addressed here and all will need to be very
carefully integrated. Within financial services, we are conscious of
the many current mechanisms for oversight, information-sharing and
collaboration, but we are also conscious of the need for better
alignment with our partners in other sectors. We look forward to
further work in this area of integration and harmonization, at both the
legislative and implementation stages.
We also believe that two areas mentioned in this section--fostering
the development of essential technologies, and cooperation with
international partners--merit considerable investment. As DHS and NIST
pursue their research and development agenda, and as the administration
pursues its recently announced International Strategy for Cyberspace,
we hope to see substantial resource commitments and advances in these
areas.
Regulatory Framework
We support all of the purposes of this section, including,
especially: The consultation among sector-specific agencies,
regulators, and infrastructure experts; and the balancing of
efficiency, innovation, security, and privacy. We recognize that giving
DHS a window into financial services' cybersecurity risks, plans, and
incident-specific information is an important element of building a
comprehensive solution. Reconciling all of these elements--Treasury and
our regulators' sector-specific roles, DHS' integration role, and the
dual objectives of flexibility and security--will be critically
important if we are to capitalize on existing oversight, avoid
duplication, and avoid the hazards of public disclosures of sensitive
information.
Federal Information Security Policies
We are encouraged by the proposal of a comprehensive framework for
security within Federal systems. As institutions report more and more
sensitive personal and financial data to regulators (and directly and
indirectly to DHS), it is critically important that this data be
appropriately safeguarded. Protecting this data, modeling best
practices, and using Federal procurement policies to expand the market
for secure products, are all good motivations for adopting these
proposed mandates.
Personnel Authorities
Because we recognize how difficult it is to recruit the most
talented cybersecurity professionals, we support the expanded
authorities articulated in this section. We particularly support
reactivating and streamlining the program for exchanging public sector
and private sector experts.
Data Center Locations
Consistent with our view of financial services as a National
market, we support the presumption that data centers should be allowed
to serve multiple geographies. We encourage Congress to consider
extending this logic for interstate data centers to the international
level, while recognizing that the owners, operators, and clients of
specific facilities and cloud networks must continue to be held
accountable for their security, resiliency, and recoverability of
customer data, regardless of the servers' geographic location or
dispersion.
h.r. 174
We share the overall objective of H.R. 174, the Homeland Security
Cyber and Physical Infrastructure Protection Act of 2011, and we
support many of its specific provisions. Listed below are a few
comments and questions that we commend to the committee as it considers
this bill and the overall issue of cybersecurity policy.
By establishing an Office of Cybersecurity and Communications
within DHS, and vesting it with the authority to establish and enforce
requirements across sectors, the bill provides for the comprehensive
treatment of cybersecurity that we have endorsed above. It offers two
options for enlisting sector-specific agencies and primary regulatory
authorities in the effort:
Delegation of authorities and responsibilities.--The Director of
the Office is given the option to delegate authority to the
sector-specific agencies and authorities. We think it is
appropriate to invest the Director with this option, much as
the administration's proposal has invested it in the Secretary
of the Department of Homeland Security and the Director of the
Office of Management and Budget.
However, given the inherent uncertainties in how this option might
be exercised, we do not believe this should be the sole
mechanism for employing sector-specific expertise and
authority.
Oversight through sector-specific agencies and authorities.--
Throughout the bill, DHS is instructed to consult with its
sector-specific partners, have private entities submit
information to them, and operate under their guidance. This
approach--with DHS setting ecosystem-level standards and sector
partners applying them as intermediaries--will reduce the
confusion and fragmentation that otherwise could occur in a
dual reporting system. We believe that financial institutions
will prefer to have their primary regulators continue to serve
as their direct supervisor on these issues, even if the
Congress determines that some requirements warrant
standardization. We believe that this approach merits
consideration, along with the standard-with-exceptions and
autonomy-with-aggregation approaches discussed in connection
with the administration's proposal.
We appreciate the bill's focus on risk-based, performance-based
regulations, rather than prescribed measures. As more detail is
developed around this approach, at both the legislative and regulatory
stages, we believe it may obviate any need for the more prescriptive
International Organization for Standardization and the International
Electrotechnical Commission standard 15408 (ISO/IEC 15408).
We appreciate the bill's commitment to sharing relevant information
to the maximum extent possible, and its designation of private-sector
submissions as sensitive security information requiring commensurate
safeguards. If other Federal Authorities are actively involved in this
process--consulting on threats, vulnerabilities, and consequences, or
as members of the interagency working group--we ask that the same
information-sharing objectives and protections apply. As the central
Department in this process, we see DHS as providing a very valuable
contribution by aggregating, analyzing, and disseminating this cross-
sector information. We encourage the committee, and ultimately DHS, to
leverage the ISACs as a key channel for these communications. We also
view research and development as a high value-added opportunity, and
appreciate the bill's attention to this function and enumeration of a
potential research agenda.
We think two of the definitions articulated in the bill are
particularly important, and therefore warrant close consideration.
First, the characterization of Covered Critical Infrastructure as
systems and assets diverges from the entity-level approach historically
applied in the financial services sector. Whether the systems-and-
assets or entity-level approach is selected, we urge the Congress to
include in Covered Critical Infrastructure not only the core of the
critical infrastructures, but also their mission-critical service
providers. In financial services, both the operational reality and the
regulatory approach require that oversight and other controls extend
well beyond the institution.
Second, because the definition of Cyber Incident drives reporting
and response protocols, we see it as a key threshold. The current
definition, as an occurrence that jeopardizes security, may be
interpreted very broadly and, without further detail, may set reporting
and response thresholds lower than necessary.
conclusion
We very much appreciate the committee's interest in the important
topic of cybersecurity, and particularly in the role DHS plays in this
element of critical infrastructure protection. Because The Financial
Services Roundtable is fully committed to enhancing cybersecurity:
We will continue to strengthen security with our members and
partners,
We will help answer this question of integrating DHS'
ecosystem-level program and the financial authorities' sector-
specific efforts,
And we will work to pass and implement the administration's
cybersecurity proposal.
Thank you very much for your time. I would be happy to answer any
questions you might have.
Financial Services Cybersecurity Policy Recommendations
financial services sector coordinating council--april 15, 2011
Policy Approaches the FSSCC Supports
Federal leadership on a National cybersecurity framework,
implemented with the active involvement, judgment, and discretion of
Treasury and the other sector-specific agencies (SSAs).
Commitment to two-way public/private information-sharing,
leveraging the Information Sharing and Analysis Centers (ISACs), the
US-CERT, safe harbors, clearances, and confidentiality guarantees. This
must include sharing of actionable and timely information.
Support focused efforts to address critical interdependencies such
as our sector's reliance on telecommunications, information technology,
energy, and transportation sectors. Continue to leverage and expand on
existing mechanisms (e.g., NSTAC, NIAC, PCIS).
Involvement of Treasury and other SSAs in cyber emergencies.
Federal cybersecurity supply chain management and promotion of
cybersecurity as a priority in Federal procurement.
Public education and awareness campaigns to promote safe computing
practices.
Attention to international collaboration and accountability in law
enforcement, standards, and regulation/supervision.
Increased funding of applied research and collaboration with
Government research agencies on authentication, access control,
identity management, attribution, social engineering, data-centric
solutions, and other cybersecurity issues.
Increased funding for law enforcement at the international,
National, State, and local levels and enhanced collaboration with
financial institutions, service providers, and others that are critical
to investigating cyber crimes and creating a better deterrent.
Heightened attention to ICANN and other international internet
governance bodies to enhance security and privacy protection.
Strengthening of Government-issued credentials (e.g. birth
certificates, driver's licenses, and passports) that serve as
foundation documents for private sector identity management systems.
Enhanced supervision of service providers on whom financial
institutions depend (e.g. hardware and software providers, carriers,
and internet service providers).
Recognize the role of Federal financial regulators in issuing
regulations and supervisory guidance on security, privacy protection,
business continuity, and vendor management for financial institutions
and for many of the largest service providers.
Policy Approaches the FSSCC Opposes
Detailed, static cybersecurity standards defined and maintained by
Federal agencies in competition with existing, private, standard-
setting organizations.
Establishment of vulnerability, breach, and threat clearinghouses,
unless security and confidentiality concerns can be definitively
addressed.
Sweeping new authority for Executive Branch to remove access to the
internet and other telecommunications networks without clarifying how,
when, and to what extent this would be applied to critical
infrastructure.
Mr. Lungren. I thank you, Mr. Williams.
Now Mr. Clinton.
STATEMENT OF LARRY CLINTON, PRESIDENT, INTERNET SECURITY
ALLIANCE
Mr. Clinton. Thank you, Mr. Chairman, Ms. Clarke, Members
of the committee. I appreciate your inviting the Internet
Security Alliance to this hearing to examine the
administration's legislative proposal.
Since ISA represents primarily companies that represent
critical infrastructure, I am going to confine my remarks to
the regulatory aspects and proposals in the administration's
plan.
The Internet Security Alliance is a multi-sector trade
organization focused exclusively on cybersecurity. We were
formed in 2000. That is nearly 2 years before the events of 9/
11, 4 years before DHS was created, 6 years before DHS created
a cyber assistant secretary, 7 years before they filled that
position, 9 years before the President appointed a cyber czar,
and 11 years before the President sent a legislative proposal
on cybersecurity to the Congress. For more than a decade, the
private sector has been leading the fight to improve
cyberspace.
During this time, we have testified several times before
Congress, constantly urging, even begging, Congress and the
administration to take a more active role in addressing our
cyber threat. There may be some in the private sector who think
that the Government should take a hands-off role in this
regard. ISA is not among them.
As the Chairman pointed out, the ISA has proposed its own
market-based system for improving our cybersecurity system, the
``Cyber Security Social Contract,'' which was cited early and
often in the President's Cyberspace Policy Review. We are not
alone. Earlier this year, several of the major organizations
that represent industry in this space--BSA, CDT, TechAmerica,
Chamber of Commerce, and the ISA--banded together to present a
detailed white paper of policy proposals for improving our
Nation's cybersecurity.
With regard to the administration's position, we find the
proposal is both too broad and too Government-centric. Although
it has been suggested that the intent of the administration's
proposal is to cover core infrastructure, we find a reading of
the legislative language rates it as far more extensive.
While there are provisions in the proposal calling for
collaboration with industry, we don't need an act of Congress
for that sort of collaboration, and the collaboration always
ends with Government fiat. For example, Section 7 requires CEOs
to certify that they are in compliance with plans required
under Section 8 and empowers the Secretary to review any
entity's plan. If DHS finds the plan wanting for some reason,
they are empowered to, ``take any action the Secretary deems
appropriate.''
In addition, paragraph 4 empowers the Secretary to evaluate
the frameworks created through various discussions with the
private sector. However, should DHS decide that the standard
frameworks don't meet their own criteria, they are empowered to
adopt their own criteria and force the companies to choose
those.
Government does not have all the answers, and it will not
be the best judge of how to manage private systems. Altering
our strategy of the public-private partnership to give the
Federal Government final say over how private companies manage
their systems will be costly, inefficient, and ineffectual.
Moreover, creating this regulatory role for DHS will
fundamentally alter the nature of the relationship between
Government and the private sector by replacing a voluntary
relationship built on collaboration with an adversarial
relationship based on regulatory mandates, reports, and
compliance. As the research I cite in my written testimony
shows, a security system based on that reactive model will be
less effective and sustainable.
Now, there is a lot we can do to improve our cybersecurity.
As the Chairman pointed out, we need to alter the economic
balance with regard to the incentives dealing with
cybersecurity. Our testimony, as well as the multi-trade
association paper, points out that there is a great deal
Congress could do to provide incentives at no cost to the
Government which will lead to the adoption of best practices
which a range of studies have indicated can stop between 80 and
94 percent of cyber attacks.
There is another area of cyber attack, many of which
Melissa mentioned earlier on, known as the APT, ultra-
sophisticated sorts of attack, that are going to require an
entirely different strategy. But we do have things in place to
deal with that also.
With regard to the administration's proposal, however, we
find that the mandatory reporting that they use will diminish
motivation for internal investigators, who may worry about
finding out material that will be harmful to their company. It
will add to the ultimate cost of detection tools and services,
making companies more reluctant to spend money on them.
Moreover, we find the evaluation program that is proposed
by the administration's proposal to be anti-security. One of
the things that everybody agrees on in this space is that we
don't have enough cybersecurity professionals. This proposal
requires virtually all entities that are covered--and that
could be many, many entities--to have annual evaluations. So we
are creating an army of insiders roaming throughout the
security procedures of our most critical networks on an on-
going basis. The value that they would have in terms of
providing actual, real security is far offset by the increased
risk of having an army of poorly-trained insiders going through
our security.
We feel it will be far more preferable for Congress to work
with DHS and the rest of the administration to create a system
where there are market incentives so that organizations will
seek to alter the balance with regard to security return on
investment--invest appropriately so that they can have
improvements in their own security and our Nation's security.
Thank you.
[The statement of Mr. Clinton follows:]
Prepared Statement of Larry Clinton
June 24, 2011
i. introduction
Good morning Mr. Chairman, and thank you for inviting the Internet
Security Alliance to testify before the Cybersecurity, Infrastructure
Protection, and Security Technologies Subcommittee.
The Internet Security Alliance is a multi-sector trade association
that develops best practices and standards, along with technological,
economic, and public policy services focused exclusively on
cybersecurity.
ISA was founded and fully funded by a group of private sector
entities in 2000. That's nearly 2 years before the tragic events of 9
///11, 4 years before Congress created DHS, 6 years before DHS created its
first cybersecurity assistant secretary, 7 years before they filled
that position, 9 years before the President appointed his first ``cyber
czar'' and 11 years before the President presented his first set of
legislative proposals on cybersecurity to the Congress.
For more than a decade, the private sector has been taking a
leadership role in the fight to secure cyber space. That is one reason
we were delighted when President Obama addressed this issue from the
White House and published the Cyberspace Policy Review shortly after
taking office--an enlightened document based on an extensive and wide-
ranging study by staff of the National Security Council.
ii. the private sector has been aggressively attempting to utilize the
public-private partnership to enhance our cybersecurity
Over the past decade, ISA has testified approximately a dozen times
before various Congressional committees constantly urging, even
pleading, for the Government to take more aggressive steps to enhance
our Nation's cybersecurity. There may be some in the private sector
that have suggested a hands-off role for the Government in this space,
but ISA is not one of them.
And, we are not alone. When legislation began heating up in the
last Congress we heard reports from policymakers that there were so
many private-sector entities that were interested in the subject that
is was becoming difficult for our Government partners to achieve
clarity as to where the private sector stood on the issue.
As a result, several of the major associations involved in this
debate banded together and worked over a period of 6 months to create a
detailed--26-page--white paper specifying our overall approach to
cybersecurity and providing detailed policy recommendations.
This unique coalition, which included the Internet Security
Alliance, the Business Software Alliance, the Center for Democracy and
Technology, Tech America and the U.S. Chamber of Commerce is noteworthy
for several reasons.
First, is the obvious size of the coalition, covering literally
tens of thousands of companies. Second, is the breadth of the
coalition. In the cybersecurity field, the ``partisan divide'' is
generally between the providers of technology and the users of
technology. This coalition included both. In addition, the civil
liberties community is represented by the most active such organization
in this space, CDT.
Finally, there is the depth of the coalition. It is not uncommon to
see a coalition of this size in the District of Columbia; however, they
are usually brought together on a 1- or 2-page letter. In this case, we
have produced an extended, and we think a cutting-edge, detailed policy
paper that analyzes a wide range of issues in the cybersecurity space
and proposes specific policies--not just broad principles.
Moreover, we sought, as much as possible to be open with our
Government partners. We took as our starting points the official
publications produced by our Government partners: the National
Infrastructure Protection Plan (NIPP) and the Cyberspace Policy Review
released by President Obama in May of 2009. Central to both these
documents is the need for the Government to work in partnership with
the private sector.
This realization has nothing to do with politics. It is based on
the fact that in cyber conflicts, it is the private sector that is most
likely to be on the front lines and it is the networks owned and
operated by the private sector that provide the critical
infrastructure--both the regulated and non-regulated ones--upon which
any modern nation relies.
Government does not have all the answers and often will not be the
best judge of how to manage private systems. Altering our strategy to
give the Federal Government final say over how private companies manage
their systems will be costly, inefficient, and ineffectual.
Cybersecurity must be achieved through a true partnership between the
public and private sectors. We specifically endorsed this foundation as
embraced in these documents:
``The current critical infrastructure protection partnership is sound,
the framework is widely accepted, and the construct is one in which
both Government and industry are heavily invested. The current
partnership model has accomplished a great deal. However, an effective
and sustainable system of cybersecurity requires a fuller
implementation of the voluntary industry-government partnership
originally described in the NIPP. Abandoning the core tenets of the
model in favor of a more Government-centric set of mandates would be
counterproductive to both our economic and National security. Rather
than creating a new mechanism to accommodate the public-private
partnership, Government and industry need to continue to develop and
enhance the existing one.''\1\
---------------------------------------------------------------------------
\1\ Business Software Alliance, Center for Democracy & Technology,
U.S. Chamber of Commerce, Internet Security Alliance, TechAmerica;
Improving our Nation's Cybersecurity through the Public-Private
Partnership: A White Paper; March 2011.
In an attempt to develop our own policy proposals via the
established partnership model, we not only notified the White House of
our intent to create the industry White Paper, but reached out to them
on a regular basis to keep them informed of our progress. We discussed
the work at the forums established under the NIPP, such as the IT
Sector Coordinating Council meetings, which are regularly attended by
DHS staff. When the paper was completed, well prior to release, we sent
a full copy to the White House for their review and comment. We
requested, and eventually received, a 1-hour meeting at the White House
to brief them on our proposals and requested on-going interaction so
that we could, as partners, come to a common ground on the way forward.
Unfortunately, no subsequent meetings were scheduled and we were never
briefed on the White House's own--substantially different--approach
until it was released and sent to the Congress.
iii. we have the tools to stop basic attacks
The committee is aware of numerous and varied cyber attacks. Indeed
the internet is under attack all day, every day, and while we
successfully deal with the vast majority of the attacks, we also must
aggressively improve our cybersecurity.
However, not all attacks are the same. Cyber attacks can of course
be segmented many ways, but given the shortage of time, we can create
two broad categories; one of basic attacks (which can be extremely
damaging) and one of very sophisticated attacks.
Most cyber attacks fall into the first--the basic--category.
Although these attacks can be devastating from many different
perspectives, they also are largely preventable.
Several different sources including Government, industry, and
independent evaluators have concluded that the vast majority of these
attacks--between 80% and 90%--could be prevented or successfully
mitigated simply by adopting best practices and standards that already
exist. Among the sources who have reported this finding we can list the
CIA, the NSA, PricewaterhouseCoopers, and CIO Magazine.
Most recently, a comprehensive study jointly conducted by the U.S.
Secret Service and Verizon included a forensic analysis of hundreds of
breaches and literally thousands of data points and concluded that 94%
of these, otherwise successful, cyber attacks could have been
successful managed simply by employing existing standards and
practices.
iv. why are we not stopping the basic attacks?
Cost.
Some have suggested that the market has failed to produce the
needed technology to address the cyber threat. That is not the case.
President Obama's own Cyberspace Policy Review documents the fact
that the private sector has developed many adequate mechanisms to
address our cyber insecurity but they are not being deployed: ``many
technical and network management solutions that would greatly enhance
security already exist in the marketplace but are not always used
because of cost and complexity.''\2\
---------------------------------------------------------------------------
\2\ Obama administration, Cyberspace Policy Review--Assuring a
Trusted and Resilient Information and Communications Infrastructure at
31.
---------------------------------------------------------------------------
This finding is substantiated by multiple independent surveys that
also identified cost as the biggest barrier to deploying effective
cybersecurity solutions. This research shows that although many
enterprises are investing heavily in cybersecurity, many others,
largely due to the economic downturn, are reducing their cybersecurity
investments.\3\
---------------------------------------------------------------------------
\3\ PricewaterhouseCoopers, The Global State of Information
Security, 2008. Center for Strategic & International Studies, In the
Crossfire: Critical Infrastructure in the Age of Cyber War, 2010.
---------------------------------------------------------------------------
The fact is that many companies don't see an adequate ROI to cyber
investments. This real-world problem cannot be permanently wiped away
by granting a Government department the power to mandate uneconomic
expenditures as President Obama himself pointed out last year at the
White House: ``Due to the interconnected nature of the system this lack
of uniform implementation of sound security practices both undermines
critical infrastructure and makes using traditional regulatory
mechanisms difficult to achieve security.''\4\
---------------------------------------------------------------------------
\4\ White House, Remarks by President Obama at White House Meeting
on Cyber Security, July, 2010.
---------------------------------------------------------------------------
Rather, we need to find ways to work within the partnership to
encourage firms to make investments that may go beyond their own
commercial risk management requirements for security, but might rise to
the level of a broader National interest. This principle was recognized
in the creation of the original NIPP:
``The success of the [public-private] partnership depends on
articulating the mutual benefits to Government and private sector
partners. While articulating the value proposition to the Government
typically is clear, it is often more difficult to articulate the direct
benefits of participation for the private sector . . . In assessing the
value proposition for the private sector, there is a clear National
security and homeland security interest in ensuring the collective
protection of the Nation's [critical infrastructure and key resources]
(CI/KR). Government can encourage industry to go beyond efforts already
justified by their corporate business needs to assist in broad-scale
CI/KR protection through activities such as:
``Providing owners and operators timely, analytical,
accurate, and useful information . . .
``Ensuring industry is engaged as early as possible in the
development of initiatives and policies related to [the NIPP].
``Articulating to corporate leaders . . . both the business
and National security benefits of investing in security
measures that exceed their business case.
``Creating an environment that encourages and supports
incentives for companies to voluntarily adopt widely accepted,
sound security practices.
``Providing support for research needed to enhance future
CI/KR protection efforts.''\5\
---------------------------------------------------------------------------
\5\ National Infrastructure Protection Plan, 2006 at 9.
---------------------------------------------------------------------------
The Obama ``Cyberspace Policy Review'' went even further in
suggesting this pathway by suggesting a mix of tailored incentives
including liability incentives, procurement incentives,
indemnification, and even tax incentives.
The multi-trade association White Paper continued this chorus of
support for this approach.
``One of the most immediate, pragmatic, and effective steps that
the Government could take to improve our Nation's cybersecurity would
be to implement the recommendations made in the CSPR to explore
incentives, such as liability considerations, indemnification, and tax
incentives. For example:
``Tax incentives that encourage establishing additional
cybersecurity investments, such as the R&D tax credit;
``Grant funding is used effectively in other homeland
security areas such as emergency preparedness and response.
Critical infrastructure industries can use grant funds for
research and development, to purchase equipment, and to train
personnel;
``Streamlining regulatory procedures, which would cut both
Government and industry costs;
``Updating the SAFETY Act to better appreciate the cyber
threat that has become more evident since its enactment. This
Act, which provides a mix of marketing, insurance, and
liability benefits for technologies designated or certified by
DHS, can be expanded to standards and practices as well as
technologies that protect against commercial as well as
terrorist threats;
``Liability protections or regulatory obligations (e.g., for
utilities) adjusting in numerous ways to provide incentives for
enhanced security practices, such as adoption of standards and
practices beyond what is required to meet commercial risks, or
enhanced information sharing. Liability benefits do not need to
be elevated to immunity to be attractive. Categories of
liability (e.g., punitive vs. actual damages) or burden of
proof levels (preponderance rather than clear and convincing
evidence) can be adjusted to motivate pro-security behavior
without costing taxpayer dollars; and
``Stimulating the growth of a private cyber insurance
industry that can both provide private economic incentives to
spur greater cybersecurity efforts while also creating a
private market mechanism that fosters adoption and compliance.
The Government should give consideration to implementing
reinsurance programs to help underwrite the development of
cybersecurity insurance programs. Over time, these reinsurance
programs could be phased out as insurance markets gain
experience with cybersecurity coverage.
To accommodate the needs of a wide variety of critical
infrastructures with different economic models, the public-private
partnership should develop a menu of incentives that can be tied to
voluntary adoption of widely-accepted and proven-successful security
best practices, standards, and technologies. The R&D tax credit may be
the most attractive option for an IT security vendor, while a defense
firm may be more interested in procurement options, an electric utility
in a streamlined regulatory environment, or an IT-user enterprise in an
insurance discount and risk transfer. Many of these incentives are
deployed successfully in other areas of the economy, but not yet to
cybersecurity.''\6\
---------------------------------------------------------------------------
\6\ Business Software Alliance, Center for Democracy & Technology,
U.S. Chamber of Commerce, Internet Security Alliance, TechAmerica;
Improving our Nation's Cybersecurity Through the Public-Private
Partnership: A White Paper; March 2011 at 10-11.
---------------------------------------------------------------------------
v. addressing sophisticated attacks
While most cyber attacks are fairly basic and can be stopped or
mitigated with the deployment of existing standards, practices, and
technologies which could be achieved through the use of a creative
incentive system, there are still other much more insidious and
sophisticated attacks that are not going to be stopped with best
practices.
Again, there are many ways to characterize these attacks but one
common term that has come to be used somewhat generically in the field
is the Advanced Persistent Threat (APT).
Without getting into the academic debate over what constitutes the
APT, it suffices to say these are sophisticated attacks. These are not
``hacker kids'' or kids in basements. These attacks are formulated by
highly sophisticated, well-organized, well-funded, often state-
sponsored attackers. These guys are pros. They are very good, and if
they target you or your system you can be pretty sure they will succeed
in penetrating, or ``breaching'' your system.
However, this does not mean we have no defense. Indeed, many
companies have been working for several years with some success on
mitigating APT attacks although it necessitates altering our defensive
posture from one of perimeter defense geared to stopping breaches to
internal detection and mitigation.
Again, the private sector White Paper identifies some of the
current core strategies that the Government, in collaboration with the
private sector ought to be deploying to address the APT style (ie. more
sophisticated) attacks. However, it is important to note that there is
no silver bullet to addressing these advanced threats.
The core reason we have attacks, and they will likely continue, is
that the economic incentives generally favor the attackers. Many
attacks are cheap, easy, and profitable while on the other hand, an
infinite perimeter needs defending, it is very hard to catch and
prosecute cyber attackers and it is difficult to demonstrate ROI to
things that you have prevented such as cyber attacks.
So long as our economic equation for cybersecurity remains out of
balance, we are going to continue to have attacks. This needs to be
understood not as a discrete problem for which there will be a simple
and unchanging security technology--like a seat belt or a set of gold
standard Government metrics. Rather, this is an on-going and persistent
threat that needs continuous deployment of creative strategies that
evolve with the dynamic threat.
vi. the administration's legislative proposal
Unfortunately, after waiting 2 years for the administration to
follow up on its CSPR, we received a legislative proposal produced
without coordination with the private-sector partnership the
administration itself had established for this purpose, and which:
Fails to follow up on the promise of earlier work by this
and the previous administration;
Does not address the core economics issues which drive our
lack of cyber insecurity;
Would create an extensive new bureaucracy that will not
address the persistent cyber threats we face; and
Could add significant new threats that are not justified by
the dubious benefits of the unbounded intrusions into our most
critical infrastructure.
Since ISA works primarily with major entities from most for our
Nation's critical infrastructure, we will focus our testimony to
Section 3 of the President's proposal, which establishes a new and
extensive regulatory structure over the private sector.
vii. the administration's legislative proposal fundamentally alters the
public-private partnership
When he released the Cyberspace Policy Review in 2009 President
Obama himself said:
``Let me be very clear: My administration will not dictate security
standards for private companies. On the contrary we will collaborate
with industry to find technology solutions that ensure our security and
promote prosperity.''\7\
---------------------------------------------------------------------------
\7\ President Barack Obama, Release of the Cyberspace Policy
Review, May 29, 2009.
---------------------------------------------------------------------------
Unlike the rigorous and open process the Obama administration
conducted in developing the Cyberspace Policy Review, the current
legislative proposal was not developed in any way by ``collaboration
with industry to find technology solutions.''
ISA participates in numerous bodies set up under the NIPP to
facilitate this sort of coordination including the Sector Coordinating
Councils, the Cross Sector Cybersecurity Working Group, the Critical
Infrastructure Partnership Advisory Council (CIPAC) and the Software
Assurance Forum. Despite repeated requests for the administration to
engage with these bodies, designated by them for collaboration to
develop solutions, there were no discussions at even a conceptual level
about this proposal which would, if enacted, fundamentally alter the
long-standing relationship.
Had the administration used the bodies designated for this sort of
interaction, I believe the proposal would be both substantively
stronger and politically more practical.
Notwithstanding the process, the Centerpiece of the proposal--the
establishment of an unbounded regulatory structured for the Department
of Homeland Security--is obviously directly at odds with what the
President pledged when he released the Cyberspace Policy Review 2 years
ago.
Obviously it will be the the committee and the Congress' decision
whether to follow this new Government-centric approach, but there
should be clarity at the very least that by establishing a broad
regulatory framework, as this proposal does, it will fundamentally
alter the nature of the relationship between the Government and private
sector.
It's often said that to a hammer, everything looks like a nail. And
prisoners and prison guards do not have a partnership. One body is
mandated to do what the other entity directs. While there is a fair
amount of verbiage in the administration's proposal about working with
the private sector, as we will discuss shortly, at the end of the day,
this legislative proposal will allow DHS to regulate pretty much any
entity it elects to regulate and mandate whatever DHS elects ought to
be mandated.
Some may argue that such a system of regulatory mandates will
finally solve our cybersecurity problem; however, there is no evidence
that this will be the case. Indeed, the academic research on motivating
investment in information security specifically points in the opposite
direction indicating that ``proactive'' investments motivated by market
incentives are more effective than reactive (prompted by regulation)
are.
A new study released from Dartmouth College earlier this month
documents this finding, ``Proactive investments are more effective at
reducing security failures than reactive investments. When proactive
investments are forced by an external requirement, the effect of the
proactive investment is diminished . . . our results show that learning
by doing through proactive security investments relies on economic
incentives whereas unilaterally mandated procedures do not have any
economic incentive . . . Government requirements simply focus attention
on the problem area rather than discovery and learning by doing . . .
external pressure does not have significant social incentives.''\8\
---------------------------------------------------------------------------
\8\ Kwon, Juhee, and Johnson, Eric; An Organizational Learning
Perspective on Proactive vs. Reactive Investment in Information
Security. Dartmouth College, NH. June 2011 at 18.
---------------------------------------------------------------------------
viii. the administration's legislative proposal is not supported by
research or precedent
Research \9\ has consistently shown that the single biggest barrier
to enhancing the cybersecurity of our Nation's critical infrastructure
is economic. As previously mentioned, the National Infrastructure
Protection Plan (NIPP)\10\ identified the need for Government to create
a value proposition for industry to make investments in cybersecurity
that are not justified by their business needs, but may be required for
overall National security. In fact, the Cyberspace Policy Review
specifically advocated the development of proactive market incentives
such as procurement, tax, and liability to incentivize additional
cybersecurity investments.\11\
---------------------------------------------------------------------------
\9\ PricewaterhouseCoopers, The Global State of Information
Security, 2008. Center for Strategic & International Studies, In the
Crossfire: Critical Infrastructure in the Age of Cyber War, 2010.
\10\ The National Infrastructure Protection Plan (NIPP) is
available at http://www.dhs.gov/files/programs/editorial_0827.shtm#0.
\11\ Executive Office of the President, Cyberspace Policy Review--
Assuring a Trusted and Resilient Information and Communications
Infrastructure, May 2009.
---------------------------------------------------------------------------
However, the administration's legislative proposal does not follow
through on any of these policy commitments.
Instead the administration's current legislative proposal relies
primarily on ``disclosure'' as a market incentive, to hoping that
reaction to such a public disclosure will generate increased
cybersecurity investment. While at one point this may have made sense,
it is not likely to be helpful when addressing the current attacks we
face.
ix. the focus on disclosure of breaches is outdated
Most cyber attack disclosure requirements are founded on
misconceptions about what it is companies have available to disclose.
Most successful modern cyber attacks go undetected. Furthermore, cyber
intrusions and malware, as they become more sophisticated and more
damaging, become increasingly difficult to detect. The tools and
services for detecting them are very expensive, and the evidence for
their presence is often very ambiguous.
The fact that the proposed legislation and the discussions that
surround it are constantly referring to ``breaches'' shows how rapidly
policy in this field becomes dated. ``Breaches'' were the big
cybersecurity concern of the last few years, but they are not the big
cybersecurity concern of the era that began with Stuxnet. What's more,
the very term ``breaches'' suggests that the remedy to cyber attacks is
perimeter defense--guarding the organization's information border
against forces attempting to penetrate, or ``breach'' it. This is a way
of thinking about cybersecurity that many of the foremost cybersecurity
experts have been arguing is obsolete for half-dozen years now. ISA
presented this finding to the Obama administration which cited the
study in their Cyberspace Policy Review and published it on the White
House website, but did not reference it in their own legislative
proposal.
In fact, most companies are unable to tell whether they have been
the victim of a successful cyber attack unless they make a special
effort to investigate, spend additional resources on the effort, and
have the necessary skills and tools already on hand. The initial signs
that need to be pursued in order to discover a skilled cyber attack are
hard to define, constantly changing, and often very subtle and thus
unsuitable for the annual evaluation procedure the administration
proposes to rely on. Uncovering a highly-skilled cyber attack is
currently much more of an art than a science. It can require intuition,
creativity, and a very high degree of motivation.
x. the administration's proposal creates the wrong incentives
Mandatory disclosure punishes companies that are good at detecting
intrusions and malware. It creates an incentive not to know, so that
there is no obligation to report. It diminishes the motivation of
internal investigators, who may worry that finding out exactly what
happened may do their company more harm than good. It adds to the
ultimate costs of detection tools and services, making companies more
reluctant to spend money on them.
Requiring companies to disclose their cybersecurity plans and
certifications is, if anything, even more likely to have unintended
consequences than requiring disclosures of successful cyber attacks.
The kinds of language and administrative formulas that would be adopted
to comply with such requirements would almost certainly have little to
do with real cybersecurity. This is partly because the field is
developing so rapidly that by the time cybersecurity plans were
recognized as fulfilling administrative expectations, they would
already be obsolete. There is also no way to tell at the level of a
``general plan'' whether the cybersecurity measures involved would be
doing any good or not. The consequence of disclosing such plans would
be another, costly level of administrative bureaucracy and auditors
that would probably only be getting it the way of good security.
xi. administration's proposed language provides dhs with unfettered and
unjustified authority over private systems
Although it has been suggested that the intent of this legislation
is to cover only the most critical ``core'' infrastructure, a careful
reading of the legislative language indicates that it provides
essentially unfettered authority to DHS to mandate technical standards
for almost any aspect of the private sector.
Sec. 3 of the Regulatory Framework for Covered Critical
Infrastructure lists a full page of requirements to be met before an
entity is subject to these, as yet unspecified, Federal mandates.
However, when reading through them, they don't provide any limit on
the Secretary's authority to designate any enterprise as a so-called
``covered critical infrastructure'' and thus subject to DHS mandates.
It's easiest to analyze the impact of the sections if we review
them in reverse order.
Subsection D states that being named on the list as a covered
critical infrastructure under this section ``shall be considered a
final action for purposes of judicial review.''
Subsection C lists a variety of criteria to be placed on a ``risk-
based tier,'' but criteria No. 4 is ``such other factors as the
Secretary deems appropriate,'' which means the Secretary can place any
entity on any tier for any reason he or she wants to.
Subsection B, which lists only 2 criteria for inclusion. One
criterion is that the entity or system ``is dependent on information
infrastructure to operate.''
Since virtually all modern systems that are reliant on some form of
information infrastructure to operate, those criteria are all-
encompassing.
That leaves us only with the criteria listed section B1, which is
that incapacity or disruption of the reliable operation of the system
would have a ``debilitating effect on National security, National
economy, or National public health or safety.''
We regard ``debilitating'' as a fairly loose, and frankly weak,
criterion for conferring such broad authority to the Secretary. To
``debilitate'' simply means to weaken--it doesn't necessarily mean to
weaken a lot--just weaken. When I catch a cold I'm somewhat
debilitated--but I wouldn't want the CDC to have the power to therefore
regulate me.
According to this legislative language, if the Secretary decides,
for any reason, that the incapacity of a system might in some way
weaken our economy, security, or safety, he or she has the authority to
mandate--as a final action--whatever technical standards over their
cyber systems the Secretary desires.
For example, the recent SONY Play Station attacks reportedly will
cost more than a billion dollars in damage, which one can argue weakens
or ``debilitates'' the economy at least somewhat. Would that then make
SONY Play Station's ``covered critical infrastructures'' under this
definition? When asked that question at a recent Judiciary Committee
hearing, an administrative witness replied that that determination
would have to be made through rulemaking under the Act.
In addition, the language does not state that the debilitating
effect referred to in Sec. (b)(1) has to be from a cyber incident.
According to this legislative language, the fact that the World Trade
Center was attacked with airplanes, which obviously had a debilitating
effect on our security and economy, would be justification for DHS to
impose mandates on the cyber systems operating in the WTC, even though
they had nothing to do with the attack.
In addition, one criteria DHS will use in assigning an entity as a
covered critical infrastructure is its interconnectedness with other
infrastructures. That again allows for a tremendous expansion of
potential DHS authority.
For example, the supply chain for weapons systems can be thousands
of companies long. Obviously, interruption of the operation of these
systems for whatever reason--including non-cyber reasons--affect our
National security. So under this language, all these thousands of other
companies would be potentially subject to DHS regulation due to their
interconnection to the main weapons system project.
Moreover, under Sec. B1 of this provision, DHS will regulate
``entities'' as opposed to systems or assets. This presumably means
that an attack having a debilitating--however minor--effect on
security, economy, or health would result in the regulation of the
entire entity the system is interconnected with.
The bottom line is that this legislative proposal provides almost
unbounded discretion for DHS to classify an entity as covered critical
infrastructure and subject the entire entity to unspecified regulation.
Section 9 states specifically that ``the Secretary shall promulgate
regulations . . . to carry out the provisions of the Title.''
Section 2 states clearly that one of the purposes of the Act is to
``establish workable frameworks for implementing cybersecurity minimum
standards and practices.''
Some may ask, ``what's wrong with DHS establishing minimum
standards for industry through a rulemaking.'' The problem is it won't
work and it is substantially counterproductive.
Now, ISA is a big fan of standards and practices and we work with
many entities, including NIST and other Federal Government agencies as
well as private sector entities to create and constantly update them.
However, there is a major difference between using the existing
consensus process to develop international standards and practices and
having a Government entity determine such standards and mandate them on
the private sector.
The multi-trade association White Paper addresses this argument in
an entire section, concluding that:
``[w]e have already seen that attempts to impose Nation-specific
requirements under the auspices of security are not embraced by the
private sector or the civil liberties and human rights community for
both public policy and economic reasons. A Government-controlled system
of standards development that resides outside the existing global
regime will not be accepted. If imposed, it would quickly become a
second-tier system without widespread user or technology community
adoption, thereby fracturing the global network of networks and
weakening its security.''\12\
---------------------------------------------------------------------------
\12\ Business Software Alliance, Center for Democracy & Technology,
U.S. Chamber of Commerce, Internet Security Alliance, TechAmerica;
Improving our Nation's Cybersecurity Through the Public-Private
Partnership: A White Paper; March 2011 at p. 8.
---------------------------------------------------------------------------
Again, although there is a great deal of verbiage discussing how
the Government will work with the private sector, the bottom line is
that this legislative proposal consistently gives DHS massive new
regulatory authority.
Section 7 requires CEOs to certify that they are in compliance with
the plans required under the Act. Although there is substantial
verbiage suggesting that DHS will work with the covered entities in
creating these plans, Section 8 empowers the Secretary to review any
entity's plan, and if DHS finds the plan wanting for some reason, they
are empowered to ``take such action as the Secretary deems
appropriate.'' In addition, paragraph 4 empowers the Secretary to
evaluate the frameworks created through various discussions with the
private sector. However, should DHS determine that the standardized
frameworks don't meet their criteria, they are empowered to adopt their
own framework to meet their criteria, and, thus, the DHS framework
would be what a covered entity would be required to implement and
certify.
xii. the administration's proposal for evaluation is anti-security
Under this proposal, an apparently enormous range of companies
would be required to construct plans for cybersecurity and plans and be
required to hire Federally-approved ``evaluators'' to review their
internal security on an annual basis. There is little if any evidence
that regulatory compliance is per se improved security. Indeed, many
report that compliance requirements distract personnel from security
work to attend to the compliance regime.
Moreover, it is acknowledged on all sides that we face a critical
shortage of qualified cybersecurity personnel and so the army of
evaluators created under this proposal will almost by definition not be
adequately trained.
The single largest vulnerability of our cyber systems comes not
from hackers using technology to break into systems but from
``insiders'' with approved access to the systems. This proposal creates
a virtual army of insiders crawling through our most critical
infrastructure's security systems on an annual basis.
The threat of introducing constant stream of new ``insiders'' into
our Nation's most critical infrastructure far outweighs the dubious
assumption that they will provide a tangible security benefit. That
does not even account for the costs industry will bear to hire these
evaluators, the cost of new manpower at DHS to comb through this
mountain of data and the potential of an ideal attack vector where all
these reports detailing our Nation's security will be stored.
xiii. the information generated by these disclosures won't enhance
security
Ironically, one of the unintended effects of more comprehensive or
stringent disclosure laws could be less information about the sort of
cyber attacks that really matter. This is because most of the mandated
disclosures would simply be noise. There would be a constant stream of
reports, based on what lawyers believe would demonstrate compliance,
while actually revealing as little as possible. This stream of reports
would obscure the attack trends that really matter, while allowing
companies to conceal events that might otherwise provoke public outcry
and more active Government intervention. As cyber attack disclosures
have become more frequent and more routine, this has already been
already happening.
The information made public by disclosure requirements is usually
not very meaningful. Most cyber attacks, even if they are successful,
do relatively little harm. They gather information that the attackers
are never able to utilize. They provide one component of a larger
attack program that never comes to fruition. In many cases, the effects
of the disclosure are considerably worse than the effects of the attack
itself. The mere fact that a company has suffered a successful attack
gives little indication of its actual losses, even if specific numbers
are mentioned. This is because there are so many factors that can
influence the scale of loss, including the wording of the disclosure
itself. Determining how much a successful cyber attack will hurt a
company is very difficult even for those who have access to all of the
details of the attack, the operations affected, and the company's
finances. For the general public, the bare facts of a successful cyber
attack are often very misleading.
The cumulative data from the cyber attacks that have so far been
publicly reported are also very misleading. Many of the biggest
reported losses of personal data were due to lost or stolen laptops.
This is not because it is the main way personal data is stolen; it is
because the loss or theft of a laptop is an unambiguous event that it
is hard not to acknowledge. Many of the other reported losses of data
have been from major defense contractors. This is not because the major
defense contractors are losing more data than other companies or than
Government departments; it is because they have the best detection
systems in place. Some of the most publicized cyber attacks have
involved Google mail. This is not because Google mail has been
compromised more than other e-mail systems; it is because Google's
business model depends more on trust and on certain types of
transparency than the business models of the other companies providing
e-mail services. Since most cyber attacks go unrecognized, the mere
fact that a cyber attack is being reported means that it is atypical.
xiv. using effective models (a) the cdc
All of this does not mean that all disclosure laws or bad or even
that the existing ones are bad. It merely points out the unintended
effects of such laws that legislators need to make an effort to avoid
in drafting additional laws. More information about cyber attacks in
general and about the degree to which individual systems and companies
are at risk is necessary for markets to take adequate account of these
things. Disclosure laws could provide some considerable benefits. But
they will not provide the intended benefits unless they take into
account how systems are monitored for attacks and what additional
information might be needed to put the attacks in context.
It is possible that the best approach might be to have the
reporting go to a special legislatively-created institution, rather
than directly to the public. This is the model used with disease
control and public health issues. With sufficiently clear instructions
as to how this institution would handle the information, its actions
could potentially be accepted by all parties. There are other ways
disclosure could be handled that would be less crude in its effects.
The point here is that any disclosure laws need to be framed with a
conscious acknowledgment of the pitfalls.
xv. effective model (b) sematech
In the 1980s, the United States also faced a technological
onslaught. During this decade, the nation of Japan began flooding the
U.S. market with computer chips, which threatened to drive U.S. chip
manufacturers out of business. Recognizing the economic and security
threat that this posed, the U.S. enacted legal measures such as the
Federal R&D tax credit and the Cooperative Research Act of 1984, which
eventually led to the private sector and U.S. Department of Defense
cooperative known as SemaTech. Within 2 years, sub-micron
architectures, advanced X-ray lithography and a number of other
critical innovations pushed U.S. chip makers back into world
leadership, and produced generation jumps in computing capabilities
just as the internet was dawning.
A similar Cybersecurity Public-Private Cooperative could be
composed of the private sector, academia, and the Government in a
minority role. This organization could be charged with improving, even
reinventing the cyber ecosystem in a more secure manner. Under this
Cooperative's umbrella, stakeholders could share information and
cybersecurity technology development to create (or fund the creation
of) more alternative networking protocols, software languages, and/or
hardware architectures that are more secure. It could also act as an
incubator for ideas to create better strategies to combat APT's and
their equivalent. It could also serve as the equivalent of an
underwriters laboratory for cybersecurity by independently assessing
best practices and standards along sliding scales. These proven
increasing levels of security, if voluntarily adopted, could then be
used to qualify enterprises for subscribing to them in return for the
incentive programs suggested earlier which will help mitigate costs
while enhancing proven security practices.
The ISA, its members and partners are aware of the need to combat
cyber threats--indeed that is why ISA was created over a decade ago.
However this must be done in collaboration with Government, not as
mandated by Government. Moreover, the solutions we derive must be both
technologically and economically practical if they are to have the
sustainable effect we require.
Mr. Lungren. Thank you very much, Mr. Clinton.
We will now go to a round of questions, 5 minutes for each
Member, and I will begin.
Ms. Hathaway, you heard Mr. Clinton's forceful testimony
there. How do you respond to that?
Let me just give a little background. I have said as a
general rule what I would like to do is to ensure that we have
a cooperative spirit between the private sector and the public
sector, No. 1. No. 2, my concern is, if we are not deft enough
in the way we have our regulatory schematic, we could--not
intended to do this--but we could have the result of stifling
creative ways of protecting against cyber attack that might
come from the private sector as we impose a Government one-
size-fits-all approach.
So I would like to see us, I guess, hit the sweet spot in
that. You have been there, you have been through these
arguments, and helped set up the contours of the debate. How do
you respond to Mr. Clinton's observation about the
administration's proposal?
Ms. Hathaway. Sir, I think that the administration's
proposal had the opportunity to engage the private sector to
inform the debate and the items within the proposal. But during
the course of their review, they did not engage the private
sector, which is why it is so important that this committee and
other committees do engage the private sector in understanding
what are the second- and third-order effects of regulation and
other market levers.
I think it will be important to take a look at both a
regulatory framework and an incentives-based framework for
research and development, for incentivizing industry to
actually get to a standard of care where we are not actually
seeing breaches on a regular basis.
Mr. Lungren. One of the concerns that I have had expressed
to me by some in the private sector--others have indicated very
strong support for the overall proposal--but one of the areas
of concern was the auditing aspect contained in the proposal,
where some suggested it was overreach.
Now, Mr. Clinton, you suggested this sort of a continual
presence there might open up the possibility of security
breaches that wouldn't otherwise exist. I suppose that is
always a balance you have to have.
How do you ensure that those that you hope are protecting
against cyber attack in the private sector, with consequences
to individuals on a more general basis, how do you ensure that
that is being done and, at the same time, don't have a heavy
hand, which may result in exposures to intrusions that you
otherwise would not have? How do you hit that balance?
Mr. Clinton. The best way to do it, I believe, Mr.
Chairman, is to make the system--to establish the system so
that the organizations want to invest in security, so that they
see it as in their own self-interest.
As I think was pointed out earlier in some of the opening
statements, what we currently have and what the National
Infrastructure Protection Plan says is that we have not
currently recognized the value proposition for industry. In
some industries, there may not be an adequate value
proposition. But there are a variety of ways that we can alter
that so that they want to invest more in cybersecurity, they
see a benefit to it.
One way----
Mr. Lungren. So they can explain to their shareholders or
justify to their shareholders and their board of directors that
it is bottom-line-effective.
Mr. Clinton. Sure. One of the ways I think you mentioned in
your opening statement is through the use of insurance. We have
not been done enough to bring the insurance industry into the
cybersecurity equation. Insurance is one of the great drivers
of pro-social behavior. We use it in health care. We use it
in--my daughter drives more carefully because she wants a
``good driver'' discount on her insurances. This affects
things. But we have not brought insurance into the
cybersecurity arena.
If we were able to motivate the greater adoption of
insurance, the insurance companies will do the evaluation for
us because their money is at risk. We can also use the
reductions in premiums to provide a motivation for the adoption
of increased best practices, just as we do when people give up
smoking to have lower insurance rates, et cetera, et cetera.
Insurance, liability reform, better use of procurement,
which has already been mentioned, streamlined regulation--these
are all things that could be offered to the private sector in
return for investing more in cybersecurity that will adhere to
their bottom line, making it so they want to do it, not because
we are making them do it, and at the same time enhance our own
Nation's cybersecurity.
Mr. Lungren. Within the administration's proposal is a
proposal for a National law on notice of breaches, which would,
as I read it, preempt the States from doing that and,
therefore, alleviate what some would say is a patchwork of
different notice requirements. On the other hand, people say
States should have the right to do that.
Does anybody on the panel have a disagreement with the
administration's approach on that?
All right.
The gentlelady from New Jersey, the Ranking Member of the
subcommittee, is recognized.
Ms. Clarke. I am from New York.
Mr. Lungren. Excuse me. New York.
Ms. Clarke. It is okay. But you know, as a New Yorker, we
have to set the record straight.
Mr. Lungren. After Mr. Pascrell yesterday indicating that
he represented the entire region, I am sorry.
Ms. Clarke. There you go. There you go.
Let me start with you, Mr. Clinton, and the whole idea of
incentivizing and the how-tos. You raised the issue of
insurance, and I want to explore that a little bit further.
Certainly, incentivizing insurance, on the surface, seems like
a proposal that perhaps could work.
What would happen if industry didn't bite or part of
industry did but the other part didn't? How do we create sort
of a uniform incentive?
Because, you know, some folks could say they want it, and
some folks could say, you know what, thanks but no thanks. Then
we are still left vulnerable, because if everyone isn't
involved, then vulnerabilities will exist.
Can you speak to that?
Mr. Clinton. Certainly, Ms. Clarke. Thank you very much.
What the ISA proposes and, frankly, what is proposed in the
multi-trade association white paper speaks exactly to your
point, which is accurate. We have a very diverse private
sector. So what we advocate is that we need to develop a menu
of incentives.
Certain incentives will be very attractive to certain
areas. So, for example, if you are in the defense industrial
base, procurement incentives are going to be particularly of
interest to you. If you are in the public utility space,
perhaps streamlining some of the regulation to make it more
cost-effective may be appropriate to you. Other sectors are
going to be interested, perhaps, in insurance. Still others
might be interested in liability reform. You have to have a
multitude of incentives, because different things will motivate
other people.
Were you also asking about how to get the insurance stuff
started?
Ms. Clarke. Well, I think my question is more to, when you
deal with things from a voluntary perspective, entities can opt
out. With cybersecurity, any opt-out equates to a
vulnerability. Any area of penetration can then have a
cascading effect. So, you know, while we want to resist the
idea of imposing anything, I am just trying to get at, you
know, how do we deal with trying to get as much coverage as we
possibly can?
I understand the menu that you have discussed. Perhaps it
is industry by industry, where we get buy-in through each
industry and its leadership, that will then cast the net that
we are looking for to close those vulnerabilities.
Would anyone else want to address that issue?
I am just trying to figure out, without imposing a
standard, if you will, how do we get everyone to see the virtue
in establishing a standard that we can hold everyone
accountable for?
Mr. Williams. Representative Clarke, if I might, I
absolutely agree with Mr. Clinton, that we should do everything
in our power to set a private-sector leadership model in this,
as we have in the past, to rely on markets wherever possible.
If the insurance and incentive models work where they work,
fantastic.
Our experience in financial services is that, with a
combination of regulatory oversight and our own business
motivations, we have done a better and better job of protecting
our sector. We have also reached out to other sectors with
uneven results. So our service providers and the sectors on
which we in a very interconnected way always depend are often
receptive to their business partners saying, ``Security is
important; we need you to invest in it,'' but not always.
That is our concern. That is our motivation for supporting
a comprehensive proposal here, is that if some opt out and they
don't happen to be in a critical tier, well, that may be
perfectly reasonable. But at least for that most critical tier,
opt-out and the possibility that at least some business
partners will just decide to go their own way and put others at
risk we think is problematic.
Ms. Clarke. Mr. Williams, let me just ask another question.
Why do you think that preemption is important? Do you think
there is a role for States in cybersecurity policy?
Mr. Williams. One way to think about the State model, as
people often describe it, is that it is a laboratory. In breach
notification and in many other areas of cybersecurity and
consumer protection, it has been a wonderful laboratory. We
have seen these breach-notification rules evolve over the last
several years with various experiments in the different States.
We believe that it is now much more mature and that now we
are ready for a National model. Those experiments have yielded
the fruit that we would expect, we have some experience now,
and we would like to see some uniformity at the National level.
The States may still very well have responsibility for, in
our case, overseeing State-chartered institutions like banks
and insurers. They may still have consumer protection
authority. But cybersecurity we think of as a National issue
where uniformity, we think, makes the most sense.
Ms. Clarke. Thank you very much, Mr. Chairman. I yield
back.
Mr. Lungren. The gentlelady yields back.
The gentleman from Texas, Mr. McCaul, is recognized for 5
minutes.
Mr. McCaul. Thank you, Mr. Chairman.
I think as you point out, Mr. Williams, I agree with the
breach-notification law. It really cries out for National
Federal law.
There are many things in the administration's proposals
that I agree with: The increased penalties for computer
hacking; the notification law; the clearer cybersecurity
authority for DHS; the FISMA reforms, which I think are
necessary. So I would have to say, overall, I think Howard
Schmidt, I think, did a pretty good job.
But the one area where I find myself in disagreement really
relates to the private sector and what role the Government
plays in regulating the private sector. I think the first
principle that we have, particularly in this area, in Congress
is to do no harm. I think we can legislate and have unintended
consequences, particularly as it applies to the private sector.
We can harden the Federal networks, and I think that is
something we are very focused on. You know, the Einstein 3--I
mean, there are a lot of things in this proposal that deal with
that. But it is really hardening the private sector and the
critical infrastructures in the private sector that I think are
the greatest challenge for us as policymakers. Ninety percent
of the critical infrastructures, up to, are really controlled
by the private sector.
So my first question is to you, Mr. Clinton. How can we
enhance that and incentivize the private sector without having
these punitive mandates?
The one thing in this proposal I disagree with is the
regulating over the private sector. Then if they are out of--I
mean, the remedy for a violation is basically what we call
``name and shame.'' You know, we will call out the company and
then publicly call out the vulnerability, which I don't think
that is very good policy, to be, you know, publicly showing
where a company is vulnerable. That just invites more mischief.
So give me your thoughts on the regulating part of this
provision, and what would you recommend?
Mr. Clinton. Well, certainly, I agree with you, Mr. McCaul,
about the disclosure aspects here. It creates a target. Not
only that, it creates an incentive for companies not to find
out things. You know, we need to incentivize people to be doing
a better job of reviewing their cyber systems.
You know, the modern cyber threat is geared around not
allowing you to know that it is there. I mean, you know, a few
years ago, cyber threats, you know, were--you had big cutesy
names like the ``Love Bug'' and ``Blaster'' and all that kind
of thing. Modern cyber threats are stealthy. They get in your
system, and the first thing they do is clean out your system,
so that when there is detection done, none of these lousy cyber
threats let you know that the really bad guy is there. They go
in your system and they hide. So it is very difficult to find
these guys.
So we want to provide incentives for people to go and look
at them. If the corporation knows that the harder they look for
a problem, the more likely are they are going to be named and
shamed for finding it, we have created exactly the wrong
incentives.
It would be much better if companies were proactively
incented in the way that I suggested with Ms. Clarke so that
they wanted to go find these things because they were going to
lower their liability, they were going to lower their insurance
rate, they were going to have a better chance at a Federal
contract, et cetera, et cetera.
The one point that I think we have to be sure, though, is
that we don't assume that there is some sort of minimum
National standard that everybody has to get to. That is not
true. The problem that we have with cybersecurity is not that
the technology is broken and so we have to bring it up to
standard; the problem with cybersecurity is that it is being
attacked from the outside. So we have to find a way to motivate
a continual investigation and innovation of mechanisms, rather
than bring people up to some sort of stable standard.
Mr. McCaul. Thank you.
My time is limited. Ms. Hathaway, I wanted to ask you a
quick question. You have a lot of expertise in these public-
private partnerships. We have had the ISACs, the information
sharing and analysis centers; have never really gotten to the
point where we want them to be. You know, when I met with some
of these firms in Silicon Valley, they talked about the
liability protections. You know, there is a FOIA exemption, or
exception, for critical infrastructures in terms of the
sharing, but there still isn't any liability protection for
them. So they are not incentivized to share information.
Can you speak to that? What would be your recommendation as
to how we can better enhance these public-private partnerships?
Ms. Hathaway. Representative McCaul, I agree that many
companies perceive that the FOIA is not strong enough if it
were actually leveraged, and, therefore, private-sector
entities are not as willing to share information.
I think that the question we need to be asking ourselves on
the Government side is, how can we share more and better
information with the private sector so they can appreciate the
threat that they are dealing with and the exposure that they
have as multinational corporations?
I think the Government does not share actionable
information with the private sector and should increase their
information-sharing mechanisms that are informed from the law
enforcement and the intelligence community.
DHS, as the forward-facing entity, needs better information
from the law enforcement and intelligence community and should
be sharing actionable information and real case studies with
the private sector of what is happening in their industry, how
certain corporations are being exposed--not necessarily naming
them, but saying company X was exposed with the following
breach and lost X quantity of confidential information. It is
only when we start using real cases and real information that
the private sector will be able to better defend itself.
Mr. McCaul. Thank you, Ms. Hathaway.
Mr. Lungren. The gentleman yields back.
The gentleman, Mr. Richmond, is recognized for 5 minutes.
Mr. Richmond. I defer to Laura my time. I think she needs
to leave.
Mr. Lungren. Oh, okay. Well, according to the rules of the
committee, it is in order of appearance. So Mr. Keating would
be next unless he allows Ms. Richardson----
Ms. Richardson. I think I was here.
Mr. Lungren. Okay. The gentlelady from California, Ms.
Richardson, is recognized.
Ms. Richardson. Thank you.
Thank you, gentlemen. That was very kind of you.
Ms. Hathaway, in your opinion, which sectors are the most
critical that we should be focusing on? We obviously can't do
everything. We are not going to have money for everywhere. In
our critical infrastructure, what would you say would be most
vulnerable?
Ms. Hathaway. Ma'am, I think that the most important
probably starts with our energy sector. Without the power, you
can't run a business and you can't sustain operations. Given
the system control vulnerabilities and in the wake of the
proliferation of Stuxnet, it is a high priority for the country
to address the vulnerabilities that are within the power
sector.
I think followed by power is telecommunications, because
without telecommunications you don't have the internet and you
don't have the ability to do e-commerce and e-business.
I would start with those two sectors.
Ms. Richardson. On a scale of 1 to 5, 5 being best
prepared, how would you rate that we would be from an energy
perspective?
Ms. Hathaway. On a scale of 1 to 5, I think that the energy
sector probably was in a better prepared state and it is now
going down the scale, as it moves more and more of its
infrastructure to an internet-based protocol and as we, the
Government, have been offering to the private sector that they
need to move more and more of their infrastructure to a smart
grid. I don't believe that a smart grid has been approached
with the security in mind first and foremost and so, therefore,
is making that infrastructure more vulnerable.
Ms. Richardson. Thank you.
Mr. Clinton, according to the White House proposal,
companies would be subject to reporting--and it was a previous
question by my colleague--would be subject to reporting
significant incidents to DHS. Do you have an objection to that?
Mr. Clinton. Well, the problem is, what is a significant
incident? As I tried to articulate in my testimony, there is
currently an opinion, a common thought in the press, anyway,
that when you have been breached, that is a significant
incident. We would probably disagree with that. In the modern
world, with modern attacks, virtually everybody gets breached.
If you are going to have some of these advanced persistent
threat guys come after you, you are going to be breached,
meaning they are going to get in your system.
That means that we have to alter the way we do defense away
from perimeter defense, keeping them out, to recognizing them
when they are in the system and mitigating the attack there. So
even though you may have been breached, that does not mean that
it is necessarily a significant incident, because, as I say,
these guys are going to get in.
If we made that the line, that you had to report the fact
that somebody successfully got into your system and then you
were subject to some of these ``name and shame'' penalties that
we discussed earlier, I think that that would be a mistake.
So it really has to do with the definition of what is a
significant incident, is where I have my problem.
Ms. Richardson. Ms. Hathaway, would you view a significant
incident being a breach, as Mr. Clinton described?
Ms. Hathaway. I think a significant incident is any time
that you lose confidential information and/or put an operation
at risk that it can no longer deliver essential services.
Ms. Richardson. Have you worked with various private
industry to define what a significant incident would be?
Ms. Hathaway. No, I have not.
Ms. Richardson. Do you have an interest in doing so?
Ms. Hathaway. I think that it is important for each sector,
whether it is the financial services, defense industrial base,
electric power, and the other 17 critical infrastructures, to
define what is a significant incident in each one of those
sectors and then define the appropriate response and mitigation
strategies.
Ms. Richardson. Okay.
Last question, for Mr. Williams: What amount of risk should
the Government be responsible for in the event of a major
cybersecurity attack in the private sector, if at all?
Mr. Williams. I think the Government is certainly
responsible for collaborating with the private sector if there
is an incident. I wouldn't say that that is the same as
accepting financial responsibility or operational
responsibility. I absolutely believe that as much as possible
of both of those need to live with those who have direct
ownership of systems and connections.
I would say that in an incident, as in a steady state, if
there is a way that we can set up the kind of voluntarily
collaboration that I think many of us support, then Government
has an obligation to participate in that process. We believe
that for DHS; we believe it for our financial regulators. We
believe that they have an opportunity to protect other sectors
when incidents like that occur. But that is very different from
accepting risk and somehow relieving others of that risk.
Ms. Richardson. Thank you.
I yield back.
Mr. Lungren. The gentlelady yields back.
The gentleman, Mr. Long, is recognized for 5 minutes.
Mr. Long. Thank you, Mr. Chairman.
Ms. Hathaway, you spoke about stiffening the penalties. To
what degree? Do you agree with the overall proposal, the
penalties that have been proposed in that? What degree do the
penalties need to be stiffened to curb some of this activity?
Ms. Hathaway. Sir, I think that it is essential that we
update the Computer Fraud and Abuse Act. Right now, we do not
have enough penalties for the breaches that are happening every
day that we read about. I think that the administration's
proposal is important.
I would take it one step further and remove the connotation
of ``protected systems.'' Protected systems are usually defined
as Government and financial institutions. I think that any
breach, regardless of where it has happened, in the private
sector, the Government, and/or in academia, should be deemed a
breach, with the same penalties.
Mr. Long. Has there been any indication that the penalties
that are there now have been effective or the increase that
they are going to in years and dollars, do you have any----
Ms. Hathaway. I believe that the stiffened and higher
penalties, if they are communicated, will start to act as a
deterrent, a domestic deterrent. I believe that, also, law
enforcement needs to have additional capacity to be able to
investigate these breaches and impose those penalties as they
find those who are committing those crimes.
Mr. Long. What percent of cyber attacks would you say are
domestic and what percent are non-domestic right now?
Ms. Hathaway. I think it would be difficult to quantify the
number of incidents and/or breaches. They are going up
exponentially every day. I think all countries are suffering
the same amount of intrusions.
Mr. Long. Okay. Thank you.
Mr. Williams, I hail from the Seventh District of Missouri,
and we had an incident there where a title company, just a
small mom-and-pop shop title company, had, I believe, $440,000
taken out of their account, their bank account, over the
weekend. This has been within the last 12 months, maybe a
little longer, 15 months, or somewhere in that neighborhood,
and had $440,000 wiped out of their account through their bank.
The Secret Service is the investigative arm that looks into
that. They have ascertained, I think, that the money first went
to Turkey, then Cyprus, ended up in Pakistan. Apparently the
hopes of getting it back are about like the hopes of me
collecting the $800 million I have been e-mailed here this
morning that is in an account in my name.
How can we protect--I mean, this is a mom-and-pop title
company. They had the financial resources and backing to be
able to go out and qualify for an SBA loan, because, as you
know, in a title company, that was not their money they were
holding. It was money they were holding for real estate
transactions to close. So they at least had the ability to go
out and borrow the $440,000, which is not a lot of consolation
to them.
But how in the world can we in Congress help the financial
services industries in this cyber attack situation?
Mr. Williams. We certainly can use some help with it. I can
tell you some of the things that we are already doing.
One of the evolutions in this whole process over the last
few years is that much of the work used to happen solely within
an institution, but now it really has to include business
clients, like the title company, in the process----
Mr. Long. And their bank.
Mr. Williams. And their bank. They absolutely need to be
cooperating so that the bank builds secure systems, the title
company secures its system and its credentials, so that they
have this collaborative arrangement where it is not entirely
within the bank's systems and the title company is not entirely
on its own in this process.
If we have more research and development, as most of these
proposals I think suggest, we will find better and better ways
to authenticate, so that if someone over a weekend has gained
the credentials of the title company, it will be harder and
harder for them to pose as a business client of the bank
without the bank being able to detect it.
Mr. Long. I don't know how we can ever get ahead of the
curve on this situation, because it seems like we are
constantly behind the curve, and the curve is moving at a rapid
pace. So if there anything, off-mike or whatever, later, if you
can get to me, as far as how Congress can help, for the entire
panel, I would appreciate it.
Mr. Williams. Yes, sir.
Mr. Long. Mr. Clinton, you made reference to the fact of
insurance two or three times. Walk me through that a little
bit. What type of insurance? What do you incentivize? The
insurance companies in this, what type of insurance are you
talking about?
Mr. Clinton. Well, there are a variety of insurance
instruments that are available--protect against breaches,
protect your liability of losses, protect your system, loss of
data. It is possible to, for example, in the example of your
title company, that they could have bought insurance----
Mr. Long. You are talking pretty much liability insurance?
Mr. Clinton. Yes, sir. The typical policies don't tend to
cover these cyber events. So there are special instruments that
are available for that.
The way that that would probably be best done--there are
two things that we propose to get that started, one of which
would be for greater information-sharing in return for some
sort of Federal benefit. One of the problems the insurance
companies have is that they don't have the actuarial data,
because companies keep that private. But we believe that,
probably, working with the Government, we could get that sort
of actuarial data. That will help to bring the rates down. If
we can get the rates brought down, then people will sell more
insurance, and we can start kind of a virtuous cycle.
The other thing, which is a much bigger idea, would be--we
have had this problem of not having enough insurance for an
important social good in the past: Crop insurance, flood
insurance, et cetera. In those instances, the Federal
Government has set up a revolving fund, and that was a better
way to manage risk.
This is one of the things I would propose that the
committee ought to look at, because right now the Federal
Government is carrying all the risk of a major cyber event. If
the East Coast goes down for 3 weeks, Congress is going to pay
for it all. That is bad risk management. You ought to be
setting up a revolving fund so that we can get some private
coverage there.
Mr. Long. Thank you.
Mr. Chairman, I have no time to yield back, but if I did,
trust me, I would.
Mr. Lungren. I was going to say, as a conservator, you are
not used to giving back something you don't have. But that is
all right. I won't interject that.
Mr. Richmond, you are recognized for at least 5 minutes.
Mr. Richmond. First of all, Mr. Chairman, let me thank you
for having the hearing, and to the Ranking Member who has been
very passionate about this issue.
The overwhelming concern that I have--and any of the
panelists can chime in--is just the country's awareness of this
as a real threat. I chaired Judiciary in the State of
Louisiana, which under our jurisdiction we had homeland
security and all of those things, and this was not an issue
that got a lot of attention, if any.
So what can we do in the importance of raising awareness of
it to help combat the threats that we have out there? Just
general public awareness, and then we can go from small
businesses to major businesses, and then we can just talk about
States, because I don't see Louisiana being prepared or being a
leader on this at the State level.
So, in any particular order. We could start with you, Ms.
Hathaway.
Ms. Hathaway. Thank you very much, sir.
I think that we do need to have a National conversation
about what is happening on our networks, and it needs to begin
really at all levels.
We need to begin the conversation about cybersecurity and
network hygiene in the K-through-12 program. As our children
are being asked to bring in thumb drives to carry their
homework back and forth between school and our home networks,
they are being used as a path to actually infect our homes that
infect our enterprises which infects our governments and
infects our banks. So we need to begin with the children.
If we then move into a university program that extends the
Information Assurance Centers of Excellence to all 50 States
and beyond 5 percent of our universities, we can start to get
to the actual practitioners of and create a stronger workforce.
If we start to have a stronger, more informed workforce on
the information security that is trained from K through 12
through university, then we start to have a better-informed
workforce and enterprises that can contribute to the National
conversation.
I would ask you, as Members, if you could go back and have
a conversation in each of one of your districts and start a
conversation in the schools and with the enterprises, because I
can guarantee you the schools have been breached or the
enterprises in your districts have been breached. You can start
a simple conversation of what it means to them and what it
means to you and how can we begin that National conversation in
every district of America.
Mr. Shannon. Yes. Thank you.
The challenge here is getting people to realize that it is
a community impact, that having one organization, one entity,
one individual compromised is really not the issue; it is when
it happens en masse. So, from CERT's experience, starting with
the Morris worm, you know, there was a realization of everyone
involved that this is a community event, it is not just their
network that has been compromised, not just their host.
So I think part of the challenge, especially when you are
looking at insurance issues and regulatory issues, is
acknowledging that community aspect. What we find is that
organizations, individuals usually are surprised when they
realize that the compromise in their system is part of an
overall industrialization of the threat and it is affecting the
whole community.
So, actually, their--putting themselves at risk, as Ms.
Hathaway mentioned, that puts everyone at risk, realizing that
we are all in this together. I think that is where the
conversation needs to lead. It is not just about your own
assets, your own data. Your vulnerabilities actually expose
everybody else.
Mr. Williams. I would have a thought or two at the family
or small-business end of the spectrum and at the more corporate
end.
At the family level, people shouldn't be worried about
advanced persistent threats or some vague notion of identity
theft. There are some very concrete things that they can be
thinking about. They can be more technology literate from the
schools at the children's level and the adults in the home.
They can be watching that their PCs and their smart phones have
antivirus protection on them, that they are well-maintained.
They can be watching their financial statements to ensure that
transactions don't appear----
Mr. Richmond. Mr. Williams, I know I am going to get cut
off in a few minutes. But if you could get me that information
or get that to the committee, I think it would be helpful.
Because a lot of us send out information to our districts all
the time, and that is something that we could put in there,
those small things to push people to do.
Before you, Mr. Clinton, I would just--you talked a little
bit about ``name and shame.'' Part of the question is the
balance between the public's right to know--because a lot of
times we, as Government, and private sector, we clash, because
the private sector would say, ``Nothing bad has happened yet.
There is no reason to act until something really, really bad
happens.'' Well, we have to take a different approach, and part
of that is to try to make sure nothing ever happens.
So how do we balance ``name and shame,'' as it is
described, with the public's right to know and the fact that
information is power, and we can prevent it that way, and not
just leaving it up the private sector until something bad
happens?
I yield back, Mr. Chairman.
Mr. Clinton. A couple things. I will try to be really
quick. Be happy to chat with you more off-line.
First of all, putting in those incentives so that we can
get to those best practices and standards that the NSA, CIA,
everybody, Secret Service, would solve 90 percent of the
problem. That is the first thing we need to do.
With regard to disclosure, ISA is very much in favor of
disclosure. But the disclosure, as I have detailed in my
testimony, the disclosures have been to be purposeful
disclosures. The public's right to be secure, I would say, is
the higher value here.
What we have proposed is, instead of having general broad
disclosure, which will go to the press, which will treat it
sensationally, will skip over the details as to whether or not
this was really harm here or not, we would propose more of a
CDC sort of model. That is where the reporting ought to be. It
should be going into entities that can understand the real
problem and can work on solving the problem so that we don't
have the losses that come out.
One of the problems here is our definitions. Think of
cybersecurity like a football game, okay? If you are the
defense in a football game, it is not a--everybody gives up
yards, right? So the fact that you have been breached, that is
not the problem. The problem is when the offense scores. So you
can have breaches that don't lead to scores.
We shouldn't be putting out publications, you know, and
having news conferences about somebody being--you know,
somebody losing just some yardage. We should confine that to
experts detailing when there has actually been losses, and then
we can deal with, you know, some sort of SEC filings that are
appropriate, which the SEC already will do.
So we are arguing for a more sophisticated form of
disclosure to deal with a more sophisticated sort of attack. We
think that that will lead to greater security, which is our
goal.
Mr. Lungren. Now, the gentleman, Mr. Marino, a great
football fan, is recognized for 5 minutes to continue the
analogy.
Mr. Marino. Thank you, Mr. Chairman.
Carrying that ball down the field on the offensive end of
things, I want to turn this conversation a little bit. We are
talking about the breaching of the systems and increasing the
penalties. But I find it ironic that we are here--and,
obviously, I am a big supporter of public hearings--but we are
here talking about security measures, which--we could have a
hacker sitting out in the audience.
So where do we draw that line between sharing public
information and not sharing it to prevent it from the hackers
getting control of it? But, by the same token, the hackers are
pretty sharp. No. 2, as far as penalty-wise, what do we do with
the 15-year-old genius who gets into the system just for fun
and causes havoc?
With those two questions, could we start with Ms. Hathaway?
My father told me ladies before gentlemen.
Ms. Hathaway. Well, let me start with the 15-year-old
genius. There are some efforts within the law enforcement
community and with the actual school districts to identify
those genius hackers. Instead of a sentencing or going to
juvenile hall, they actually start working with the law
enforcement community or get prepared to work for our
intelligence community. So they are the next-generation
workforce with the skill set that we need.
Mr. Marino. Okay. Let me interrupt just for a moment.
Primarily--and the former attorney general from California will
agree with me, I think, that the Federal system has very little
jurisdiction or, actually, maybe no ability to deal with
juveniles.
Ms. Hathaway. I understand that the law enforcement
community has been working with the high schools to actually
help identify and work with using their skill set and turning
it to good as opposed to harm.
Mr. Marino. I understand that. But how about the penalty
aspect of it? What is your position on that? Do you have a
suggestion on that?
Ms. Hathaway. I think that penalties for kids, we need to
look into, the penalty could actually be serving, you know, for
the U.S. Government or serving on behalf of the communities to
actually go out and prosecute.
Mr. Marino. Mr. Shannon.
Mr. Shannon. Could you repeat the question? I have lost the
track of what you--the first part of the question.
Mr. Marino. The two questions were: Keeping it
confidential; and how do we deal with the juveniles? Because
the Federal system is not that well-equipped to deal with
juveniles when it comes to penalties.
Mr. Shannon. Yeah, I will deal with the confidentiality
issue.
One of the great innovations of the internet is the freedom
to express yourself, the freedom to create new technical
capabilities, to innovate quickly. It is enabled by open
disclosure, open sharing of information.
Clearly, disclosing vulnerabilities is a challenge, but
when you realize that there is a threat and there is a
remediation, sharing that quickly and openly is better than
what is the alternative, remaining ignorant. Because I can
assure you that the hackers do know, and if you try and
communicate it in some out of sort of closed or secure manner,
once you get to sufficient scale, they will still know. So, you
know, there is no hiding it, in that sense.
So it is better to put the information out there, let
people be informed, and then they can make the appropriate
decision, especially when it comes to a mitigation.
Mr. Marino. Okay.
Mr. Williams.
Mr. Williams. I think, quite appropriately, most of this
conversation already occurs in confidential spheres and should
continue that way. So companies, when they contract with other
companies, will talk very explicitly about their security
posture. That has a very strong market incentive for people to
do the right thing.
In our industry, institutions talk with their regulators,
but they do that almost exclusively behind closed doors. The
kind of sharing that I think the administration proposal
contemplates would also be confidential, two-way sharing
between DHS and some of the other agencies and the companies.
There are, I think, a couple of exceptions to this idea
that there should be a cloak of confidentiality generally. One
is, if there is information that can help consumers to protect
themselves, if an individual consumer has been put at risk,
there are and should be rules to ensure that that person knows
what they need to know to protect themselves. The same at the
SEC level for investors.
Mr. Marino. Okay. All right. Thank you.
Mr. Clinton, you have 18 seconds.
Mr. Clinton. We are dealing with different levels of data,
so the sophisticates ought to be meeting amongst themselves and
sharing data and then atomizing it and then have it pushed out
to the broader community.
We have a proposal we actually started with Melissa
Hathaway a couple of years ago with DHS to do exactly that. I
would be happy to talk with you more off-line.
Mr. Marino. Thank you. Touchdown.
Thank you very much, Mr. Chairman.
Mr. Lungren. Now I will be happy to recognize the gentleman
in this Congress who probably was happier than any other Member
that Whitey Bulger got nabbed yesterday, Mr. Keating.
Mr. Keating. Happy and relieved.
You know, interestingly enough--I will just a little share
information with you--in terms of getting the word out, I was
struck by the fact that there is a group in the Boston area
where the 30 top executives, largest firms, they meet usually
annually to discuss what their biggest issue is. That could be
taxes, it could be anything; it is open-ended. They decided it
was cybersecurity. So I do think that people understand the
magnitude and the importance of this, and that is out there.
What I am struggling with is this, and I don't know if
there is an answer. Mr. Clinton started down that track, but I
would just like to ask the rest of the panel if they could help
in this regard. I am looking for something, an existing model,
public-private model, quasi-governmental model, that already is
there, may not be a perfect fit, but just to give me an idea of
where the Chairman said, the sweet spot is. We are looking for
something that is flexible enough so that regulations don't
smother the ability and provide deterrence.
But I don't agree with, you know, the CDC model approach,
that, you know, it is just out there. I think we have to more
oversight proactively on that. I don't know where that is. I
know that the ``name and shame'' issue can, I think, be
mitigated by having, you know, rankings, the way they do in
financial institutions. When they do an audit, you can have
CAMEL ratings, whatever ratings they might be--1, 2, 3, 4--and
you are in categories where, you know, companies will have some
responsibility, and insurance companies can look at that as
well.
But if you could--and I don't anticipate anyone has a
perfect fit--can you think of some existing models in other
areas? You know, Mr. Clinton has mentioned the CDC. I would
like to ask the other panelists.
Mr. Shannon. So I think there are a couple of models. There
is the automotive and airline industry that, you know, have
reporting on accidents and incidents that allows for an
appropriate oversight. So it is a more closed--the NTSB, you
know, has a closed investigation when an incident happens.
I think it is important also to look at the CDC model and
think about where it actually is appropriate and where it is
not appropriate. I mean, where it is not appropriate maybe is
nation-state threats. But certainly in terms of deal with
industrial challenges in malware and exploitation, being able
to have a better situational awareness based on the
preponderance of incidents is what is needed.
An individual, just because I got hacked, I don't know if I
am the only one in the world or whatever. But if the Government
wants to or organizations want to be able to do a broad
response, having that sort of situational awareness is
imperative. Otherwise, you don't know that there is a
challenge.
Mr. Keating. Thank you, Dr. Shannon. That is great after
something has happened, too, and that is important.
What about trying to prevent areas and to rank or to find
some kind of oversight that is not too, you know, over-
regulatory in nature?
Mr. Shannon. I will defer to my colleagues. We deal with
things when----
Mr. Keating. That is all right. Thank you.
Mr. Williams. If I might, one macro example, one micro
example.
A macro example I think is environmental protection. There
was a time when the best thinking on environmental protection
was simple command-and-control regulation. I don't think that
is the right model for us here.
But over time, environmental protection advocates realized
that industry needed to be at the table in determining what the
solution was and then also needed to be at the table in
executing it. I think that is where we are in cybersecurity. We
need to work together to figure out what the right answers are
and then to deliver them.
The micro example, just the information-sharing and
analysis center within our sector I think is a good model of
public-private collaboration. It is largely chartered and, in
many ways, supported by Government resources. It helps us
connect with other sectors. But it is a private-led, voluntary
effort that we think has brought us great progress.
Mr. Keating. Ms. Hathaway, did you have any thoughts?
Ms. Hathaway. I think that there is a lot that could be
done by turning to the internet service providers and the
telecommunications companies as the first order of warning and
defense.
Australia has adopted a code of practice or a code of
conduct where 90 percent of their telecommunications providers
have opted in, without regulation, to provide that service to
the core infrastructure. Europe, within the European Union,
have adopted Telecommunications Directive 13a, which is
regulating all of the internet service providers within all 27
countries to provide that service across their infrastructure.
I think that the United States could learn from those
different experiments and/or capabilities and understand what
the costs are to better clean and keep our infrastructure clean
and warn us of the impending threats.
Mr. Keating. Great. Thank you very much.
Mr. Lungren. I thank my fellow Members of the subcommittee
for attending.
I thank the witnesses for their valuable testimony. This
has been very, very helpful. It is the beginning of the
inquiry, in a real sense, rather than the end of it.
Members of the committee may have some additional questions
for the witnesses, and we would ask you to please respond to
those in writing. The hearing record will be held open for 10
days.
The subcommittee stands adjourned.
[Whereupon, at 11:50 a.m., the subcommittee was adjourned.]
A P P E N D I X
----------
Questions From Chairman Daniel E. Lungren for Melissa Hathaway
Question 1. From media reports, China is engaged in the most
damaging hacking campaign in history. At the same time, its primary
telecommunications equipment provider continues to gain U.S. market
share, including in the Federal market.
What solutions can the Federal Government pursue against foreign
espionage? How can the private sector protect against the threat?
Answer. The Webster's definition of espionage is, ``the practice of
spying or using spies to obtain information about the plans and
activities especially of a foreign government or a competing
company.''\1\ There is a long history of espionage and in general, it
is a globally accepted practice of intelligence collection to better
understand Government and company intentions. The National Counter
Intelligence Executive (NCIX) tracks these trends and reports to
Congress the status of foreign economic collection efforts and
industrial espionage.\2\ In its fiscal year 2008 annual report, NCIX
reported that ``foreign economic intelligence collection and industrial
espionage has continued unabated.''\3\ The newspapers highlight
everyday that companies and governments regularly face attempts by
others to gain unauthorized access through the internet to the
information technology systems by, for example, masquerading as
authorized users or through the surreptitious introduction of software.
However, this does not negate the need to limit foreign espionage that
has become increasingly more pervasive and sophisticated against our
public and private sectors. Furthermore, focusing on one opponent may
distract our industry and Government from implementing a more complete
strategy.
---------------------------------------------------------------------------
\1\ http://www.merriam-webster.com/dictionary/espionage.
\2\ Industrial espionage, which is the knowing misappropriation of
trade secrets related to or included in a product that is produced for
or placed in interstate or foreign commerce to the economic benefit of
anyone other than the owner, with the knowledge or intent that the
offense will injure the owner of that trade secret. Misappropriation
includes, but is not limited to stealing, copying, altering,
destroying, transmitting, sending, receiving, buying, possessing, or
conspiring to misappropriate trade secrets without authorization.
Industrial espionage is also criminalized under the Economic Espionage
Act.
\3\ http://www.ncix.gov/publications/reports/fecie_all/fecie_2008
///2008_FECIE_Blue.pdf.
---------------------------------------------------------------------------
Potential Solutions
The Federal Bureau of Investigation (FBI) and the
intelligence community need to better inform industry of the
threats they are facing and how they are being exploited or
penetrated. A training program to educate corporate leadership
on how to mitigate the risk of being a high-value target,
including providing them with briefings about the threat to
their industry using specific case studies, would go a long way
to reducing the number of incidents and loss of confidential
information.
DoD is proposing to amend the Defense Federal Acquisition
Regulation Supplement (DFARS) to add a new subpart and
associated contract clauses to address requirements for
safeguarding unclassified DoD information. This development is
essential because emerging, pre-classified military
technologies or commercial breakthrough technologies are
increasingly becoming the target of espionage. The proposed
DFAR changes would require industry to implement basic security
measures to increase their defenses from cyber intruders.
Engage the United States Department of State's International
Telecommunication Advisory Committee (ITAC) \4\ and the
Advisory Committee on International Communications and
Information Policy (ACICIP)\5\ to better understand predatory
trade practices in the United States and elsewhere and develop
strategies to respond to these practices in a timely manner.
Use these advisory councils and others to gain a better
understanding of what trade and economic implications are to
U.S.-based corporations if other countries impose a Committee
for Foreign Investment in the United States (CFIUS) like regime
to protect their respective National and economic security
interests.
---------------------------------------------------------------------------
\4\ The United States International Telecommunication Advisory
Committee (ITAC) advises the Department of State in the preparation of
U.S. positions for meetings of international treaty organizations,
develops and coordinates proposed contributions to international
meetings as U.S. contributions, and advises the Department on other
matters to be undertaken by the United States at these international
meetings. The international meetings addressed by the ITAC are those of
the International Telecommunication Union, the Inter-American
Telecommunication Commission (CITEL) of the Organization of American
States, the Organisation for Economic Co-operation and Development
(OECD) and the Asia-Pacific Economic Cooperation (APEC). Members of the
ITAC are drawn from the Government, network operators, service
providers, and manufacturers involved in the telecommunications sector.
\5\ The Advisory Committee on International Communications and
Information Policy (ACICIP) serves the Department of State in an
advisory capacity concerning major economic, social, and legal issues
and problems in international communications and information policy.
These issues and problems involve users and providers of information
and communication services, technology research and development,
foreign industrial and regulatory policy, the activities of
international organizations in communications and information, and
developing country interests.
---------------------------------------------------------------------------
Congress should consider updating the Economic Espionage \6\
Act of 1996. While the definition of trade secret is consistent
with the Uniform Trade Secrets Act, which states that the
information is subject to reasonable measures to preserve its
secrecy and derives independent economic value from not being
generally known to or ascertainable by the public, the
threshold for protection is too high. As such, industry is
required at the onset of the development to protect any idea as
a trade secret. Addressing the broad-based economic industrial
espionage that we are observing on our corporate networks
requires that the Government lower the threshold for a trade
secret or add a threshold around proprietary information.
---------------------------------------------------------------------------
\6\ Economic espionage, which is the knowing misappropriation of
trade secrets with the knowledge or intent that the offense will
benefit a foreign government, foreign instrumentality, or foreign
agent. Misappropriation includes, but is not limited to, stealing,
copying, altering, destroying, transmitting, sending, receiving,
buying, possessing, or conspiring to obtain trade secrets without
authorization. Section 101(a) of the Economic Espionage Act (EEA) of
1996 criminalizes economic espionage.
---------------------------------------------------------------------------
Question 2. A large issue facing appropriate risk management for
Government and critical infrastructure is supply chain risk management
since so much of our software and IT equipment is manufactured
overseas.
What's the best approach for better evaluating the security of our
IT supply chain?
Answer. The internet and the information communications
infrastructure has evolved and has been enhanced by global commercial
innovation. While the United States incubated its beginning through the
Advanced Research Projects Agency (ARPA) in the late 1960s, and helped
it flourish through Palo Alto Research Center and the companies of
silicon valley, its evolution and the attendant benefits to society
have come from many other countries and global corporations. Our
infrastructure is dependent on this global marketplace and our economy
is dependent upon this backbone remaining secure and resilient. A
broad, holistic approach to risk management is required rather than a
wholesale condemnation of off-shore development, foreign products and
services, or foreign ownership.
The best approach to securing our IT supply chain is one that is
transparent, mindful of unintended second order consequences, and aids
in decision making. We must recognize that the supply chain consists of
many phases: Design, manufacture, integrate, distribute, install and
operate, maintain, and retire--and any conversation regarding security
of the supply chain must apply to the entire lifecycle. To meet
tomorrow's threats, we must develop protection measures across the
product lifecycle and reinforce these measures through acquisition
processes and effective implementation of agency security practices.
For example, the highest risks in the supply chain are ``after build''
(e.g. install and operate and retire phases) because this is where
multiple vendors participate in the process (e.g., integrate products
with other systems, patch/update, etc.) and there are few measures to
monitor and assure integrity throughout the entire process.
To understand alternative approaches will require a partnership
with industry that assures coordination and buy-in that enables
industry to ``do the right thing'' and not be penalized in the process.
A dialogue has begun via the Open Group Trusted Technology Forum and it
enjoys international participation by governments and industry alike.
The Open Trusted Technology Provider Framework sets forth best
practices identified by a cross-industry forum which, if used by a
technology vendor, may allow a Government or commercial enterprise
customer to consider the vendor's products as more secure and trusted.
Moreover, the Cyberspace Policy Review called for the need to
``define procurement strategies through the General Services
Administration, building on work by the National Security Agency for
the Department of Defense, for commercial products and services in
order to create market incentives for security to be part of hardware
and software product designs, new security technologies, and secure
managed services.'' The efforts of the United States General Services
Administration (GSA) in working to address this requirement through the
SmartBUY blanket purchase agreement awards aimed at providing better
cybersecurity protection to Federal, State, local, and Tribal
governments should be strongly supported. Under its new Federal-wide
Situational Awareness and Incident Response (SAIR) Tier II
cybersecurity initiative, GSA will use the procurement process to help
protect our IT infrastructure from cybersecurity incidents and other
vulnerabilities, while providing maximum value for taxpayer dollars.
These two initiatives are good steps toward enhancing the security
of the supply chain while at the same time being mindful of market
forces.
Question 3. Will the administration's proposal of DHS authority
over the private sector--which envisions Federal ``framework,'' used to
develop cyber plans, and a subsequent evaluation of those plans--
provide the necessary flexibility to optimize private sector security?
Answer. Not necessarily. The legislative proposal states, ``the
owners or operators of covered critical infrastructure shall develop
cybersecurity plans that identify the measures selected by the covered
critical infrastructure to address the cybersecurity risks in a manner
that complies with the regulations promulgated, and are guided by an
applicable framework designated.''\7\ This proposal attempts to
establish a minimum standard of care and an audit and certification
function that would be similar in kind to the Securities and Exchange
Commission (SEC) requirement for attestation of material risk.
Inserting DHS into a regulator role runs the risk of diluting its
operational and policy responsibilities, which would detract from the
Nation's security posture. In May 2011, Senator Rockefeller asked the
SEC to look into corporate accountability for risk management through
the enforcement of material risk reporting.\8\ And in June 2011,
Chairman Schapiro said that the SEC would look into the matter. If
Congress believes corporations should meet such a reporting requirement
then it should turn to the SEC, which is the Executive Branch
Independent Agency responsible for this type of reporting, and not add
an additional mission responsibility to DHS.
---------------------------------------------------------------------------
\7\ The White House. Cybersecurity Legislative Package:
Cybersecurity Regulatory Framework For Covered Critical Infrastructure
Act.
\8\ Senator Rockefeller letter to SEC Chairman Mary Schapiro. 11
May 2011.
---------------------------------------------------------------------------
Question 4. Will the authorizations for DHS to ``work with'' the
Federal Acquisitions Regulatory (FAR) Council to improve supply chain
security have any practical effect?
Answer. It is unclear. Adjusting the way that the Government
procures goods and services can be a catalyst for change but may not
necessarily make a material difference in the security of the supply
chain. The key is to decide what are the measures of performance that
are desired and under what conditions? If the level of security
assurance increases, but price goes up unacceptably, is that success?
Changes to the FAR can certainly result in change to business
processes. The changes in business processes may result in increased
costs which will be passed onto the Government and other customers.
It also is important to realize that any change to the FAR may not
apply across the Federal Government. Some agencies are exempt from
these rules including: the Central Intelligence Agency, the United
States Postal Service, the Tennessee Valley Authority, the Federal
Aviation Administration, and the Bonneville Power Administration. In
these cases, the agency promulgates its own specific procurement rules.
Question 5a. The White House has directed that all Federal
Departments and Agencies move a portion of their data processing and
storage to the cloud in the coming years.
While that strategy is a good one when it comes to making the most
of Federal IT spending in these fiscally demanding times, how can the
security of the cloud be evaluate and improved to ensure that we're not
taking unnecessary risks with mission-critical data?
Answer. According to the National Institute of Standards and
Technology (NIST), ``Cloud computing is a model for enabling
convenient, on-demand network access to a shared pool of configurable
computing resources [e.g., networks, servers, storage, applications,
and services] that can be rapidly provisioned and released with minimal
management effort or service provider interaction.'' The key tenet of
the cloud is availability. But the other two cornerstones of
information security--integrity and availability--are not readily
commanded by the cloud environment. The October 2010 report from
Forrester on cloud security states that security is the single biggest
barrier to broad cloud adoption.
In December 2010, the Office of Management and Budget (OMB) issued
a report entitled the ``25 Point Plan to Reform Federal Information
Technology Management'' and in February 2011 it published another
report entitled, the ``Federal Cloud Computing Strategy,'' where it
articulated the need for: Consolidation, efficiency, and reduction in
IT spend. The second report directed each department and agency to
identify three ``must move'' services within 3 months, and move one of
those services to the cloud within 12 months and the remaining two
within 18 months. Most departments and agencies are looking to move
email to the cloud as their first project.
The GSA is developing a contract vehicle to service agency needs
for cloud computing, entitled Federal Risk and Authorization Management
Program (FedRAMP). Many within industry are raising substantive
concerns with the proposed controls and specifications as being too
difficult and costly, and that they potentially could prevent vendors
from being able to move agency computing operations to the cloud by the
deadline. Any cloud environment that is to be used to process
Government workloads must be able, at a minimum, to demonstrate that it
provides the same level of security (as defined in the question) as a
traditional system. Currently, this is demonstrated via a Federal
Information Security Management Act (FISMA) certification and
accreditation (C&A) process, which process has been roundly criticized
as a compliance-based framework focused upon a snap-shot in time. While
one can argue that a cloud computing environment can be made more
secure than a traditional one by leveraging certain aspects and
features of virtualization and other enabling cloud technologies, the
security ecosystem (technologies, control frameworks, audit procedures,
threat models, etc.) must account for the unique attributes and
vulnerabilities of cloud computing to be relevant.
Having said that, several large-scale efforts are in progress, in
both Government and industry, to rigorously measure risk related to
cloud computing implementation. Among these are: (1) The ``Proposed
Security Assessment & Authorization for U.S. Government Cloud
Computing'', drafted and released for comment and public input jointly
by National Institute for Standards and Technologies (NIST), GSA, the
Federal CIO Council, and some of its subordinate working bodies; (2)
the Cloud Security Alliance, an industry association centered on cloud
computing, has developed a Cloud Controls Matrix, which cross-connects
established security requirements in the Health Insurance Portability
and Accountability Act of 1996 (HIPAA), the Health Information
Technology for Economic and Clinical Health (HITECH) Act, International
Standards Organization (ISO), IEEE, NIST publications, FedRAMP, and
other sources; and (3) the Defense Science Board has launched a task
force to review cybersecurity and reliability in a digital cloud. There
is broad agreement among serious information security practitioners
that the task of defining security standards for cloud computing is a
work in progress, and several organizations have commissioned studies,
(e.g., the Intelligence and National Security Alliance (INSA) and the
Armed Forces Communications and Electronics Association (AFCEA)) now in
progress, to evaluate and report on specific aspects of the subject. In
my opinion, one of the best, objective reports that describes the
opportunities and vulnerabilities associated with cloud computing, is
one that was published in November of 2010 by the European Network and
Information Security Agency (ENISA), entitled: Cloud Computing:
Benefits, Risks, and Recommendations for Information Security.
Question 5b. How can continuous monitoring be implemented in the
cloud environment? Do you have any current examples of strong security
in the cloud?
Answer. It is important to recognize that the term ``cloud
computing'' embraces several different technical and process models,
which by their nature have highly-differentiated levels of monitoring
and control by sponsoring organizations/hosts, and concomitantly, very-
different levels of active participation by hosted entities. And when
considering these, one must keep in mind that the implementation of the
cloud is the most important aspect and no two clouds are implemented
exactly the same.
How continuous monitoring gets implemented in the cloud very much
depends on the type of cloud environment and the willingness and
capabilities of the provider to conduct continuous monitoring
activities. There are numerous technologies that exist today or that
are in development to enable the monitoring of the cloud (e.g.,
infrastructure, systems, and data). The real question is: What is being
monitored and does it actually correspond to the proper threat model?
The United States Department of State may be an example to turn to as
it has the ``first mover advantage'' for use of a secure cloud
environment. It is applying a high degree of rigor in timely scanning
and prioritized remediation through continuous monitoring--thereby
providing a more secure common baseline for all.
As such, there can be no general answer to this question. However,
certain ``private clouds'', hosted by highly-competent security
organizations and providing infrastructure, platform, and/or software
services to members of their own organization only, may be considered
highly-secure. Examples would include certain clouds developed and used
inside National intelligence agencies, hosted on-site and with access
limited to authorized employees of those organizations. In such cases,
the economic virtues of efficiency and economy of scale is use of IT
resources may accrue, but security of hosted data, participants,
infrastructure, and services are all tightly controlled.
Questions From Chairman Daniel E. Lungren for Gregory E. Shannon
Question 1a. DHS has been developing the National Cyber Incident
Response Plan, which it exercises through its bi-annual response plan.
What more should the Federal Government be doing to improve
response to cyber attacks?
Answer. Encourage more frequent agency and interagency cyber
exercises that will identify technological and procedural gaps as well
as build working relationships and trust both within and across
agencies. For any response activity to be effective the organizations
that participate in the response need regular, structured, measured
practice, weekly or monthly if possible. This practice builds common
understanding of the processes and technologies to be used as well as
builds trust among the various participants. These exercises need not
be immense/expensive; smaller-scale exercises testing various
subcomponents of a response plan on a regular basis would be valuable
and cost-effective.
Support timely access to operational situation and incident data.
The Federal Government should study the history of the PCII program and
the lessons learned to update it to be more attractive to industry.
Encourage making meaningful sets of operational data accessible to
researchers so that they can determine what data is best to share and
what prevention/response tactics are most effective.
Question 1b. Are there priorities for DHS response planning that
would be helpful to include in legislation?
Answer. Priority: How the Federal Government should engage the
private sector in a major incident--what information should agencies
provide and when will they provide it? Plans should include: How to
engage the private sector in a major incident, which entities does the
Government need cooperation from, and how is best to collaborate? This
will make the Government more predictable; allowing the private sector
to then plan appropriately.
Priority: Grant Federal CIOs more authority for protecting their
cyber infrastructures before incidents occur. It's difficult, if not
impossible, to defend that for which one had no hand in creating.
Question 2. When CMU-CERT is engaged in a response to a cyber
attack, what is the greatest difficulty of getting information from the
private sector?
Answer. There are several significant barriers to getting the
private sector to share information.
The Federal Government is frequently hard-pressed to convince the
private sector that there is real value in sharing information with
them. The perception continues to be that when industry shares
information they receive nothing (or nothing of value) in return from
the Government. CERT has been a part of successful models, such as the
work done by DC3 in the operation of the DoD-Defense Industrial Base
Collaborative Information Sharing Environment (D-CISE), whose example
could be built upon in other critical infrastructure sectors. It takes
effort to demonstrate to the private sector that the Government can be
helpful; e.g. by extracting indicators from sensitive data or by
creating the environments and the tools for cleansing data so it cannot
be attributed to its source and thereby shared with the private sector.
Additionally, the private sector has multiple concerns about the
potential adverse effects of sharing information--those barriers, such
as fines, litigation, etc. should be identified and eliminated through
incentives and safe harbors, where possible.
On the other hand, while some entities might not want to share, a
large number of companies (particular small- to medium-sized
businesses) do not have the capabilities to collect and/or analyze the
data that is necessary for their own protection much less useful to the
Government. What needs to be shared is actionable information and the
capability to successfully implement the actions must still be built.
The Government could encourage industry to develop the competency using
incentives (e.g. the Government could consider subsidizing these
competencies thru CNDSP or MSSP models).
How information comes to the Government can also have significant
impact on whether or not the Government can disseminate critical data
to prevent further impact to other entities. In many cases, the
Government knows what is happening, but effective communication of
remediation information is often limited. At times, information comes
in via reporting with so many restrictions that the Federal Government
cannot share the data. Savvy organizations are realizing that they can
``have their cake and eat it too'' by complying with the reporting
requirements but while still ensuring that no data, remediation
information, or conclusions from their incident is distributed.
Lastly, the Government, in its handling of classified information,
should examine current practices to find effective ways to separate
actionable information from classified or privileged data so that
incident data can be used to help others protect themselves.
Question 3a. How do we bridge the gap between operations and
research to transition technology in a timely and effective manner?
Answer. In order to effectively transition research for real-time
operations there must be stronger feedback mechanisms between
operations and research.
We believe CERT is a successful model that brings together
researchers and operators and could be an effective paradigm for
others. The CERT Program at the SEI, through its customer engagements
with security operations centers, network operators, vulnerability and
malicious code analysis centers, incident response teams, law
enforcement investigators, and intelligence analysts, has a first-hand
view into the state of security in our National critical information
infrastructures. This view helps us understand the security strengths
and weaknesses of fielded technology and systems, the evolving threats
and associated attack methods and tools, the effectiveness of current
security technologies and practices, and the security needs of system
operators. Empirical data from our DoD and other Government customer
engagements ensure our research and development agenda is grounded in
operational problems and realities, and we are addressing significant
problems for which effective solutions do not currently exist. This
model also creates an environment where solutions can be rapidly
deployed and prototyping with strategic customers helps set realistic
transition paths for the broader community.
The challenge in transitioning potentially important cybersecurity
innovations from small companies and startups is especially profound.
Having spent half of my (Dr. Shannon) career in such companies, I know
this challenge first-hand; it is difficult, if not impossible, to get
timely operational feedback on one's technology when dealing with
Government customers. I encourage the subcommittee to support efforts
to bring together operationally relevant data and small companies so
that: (1) Government entities can determine if there's promise in the
technology, and (2) the small company can quickly iterate and adapt to
the realities of the operational data.
The challenge is to create a continuous capability with steady
inflows of technologies, operational knowledge, and Government needs.
CERT/SEI/CMU is already doing this successfully but intermittently for
specific customers with innovations from academia. Sustaining this
activity at CERT and elsewhere and expanding it to small companies
would improve the flow of effective cybersecurity and incident response
innovations into the Government.
Question 3b. And what resources are needed?
Answer. The Government would greatly benefit from establishing and
maintaining a sustained cybersecurity and response innovation
acceleration program focused on transitioning innovations from the
private sector to the Government with subsidies for small businesses
and universities and incentives for larger businesses. This endeavor
could be funded at $4-6 million/year and would bring four essential
elements together: Unique operational data sources, private
innovations, informed scientific evaluation, and Government needs. The
goal, from first contact with a company, would be to operationally
deploy their validated innovation(s) in less than a year within some
meaningful part of the Government.
Question 4. How can we increase our confidence in the various
technical and policy solutions proposed at any point will be as
effective as promised/implied?
Answer. Encourage the use of scientifically validated metrics and
measurements in studies about proposed solutions. Too often
cybersecurity solutions proposed have been based on limited evidence
and/or scientifically unvalidated data and techniques.
The ability to measure effectiveness of technology and new policy
is an area sorely in need of research and deeply in need of funding. I
(Dr. Shannon) am truly humbled at how little that we experts say we
``know'' about cybersecurity and incident response that has actually
been scientifically validated. Research sponsors should be encouraged
to invest in ``the empirical science of cybersecurity'', including the
development of metrics and experimental methods that support
measurement of the effectiveness and cost/benefit of proposed security
solutions.
Question 5a. In your testimony you mention that the Government
should focus on three things to improve incident response capability,
information sharing, forensic analysis capability and training.
Focusing on information sharing, in your opinion does requiring
reporting improve the quality of reporting or just the quantity?
Answer. Today, such a requirement would only increase the quantity.
Per our answers to the other questions above, research into what is the
right data to share as well as cost-effective means to collect and
analyze the data will enable mandatory reporting requirements to
improve the quality of the data.
With mandatory reporting requirements should come clear guidance on
what data and associated meta-data needs to be shared; under what
circumstances; ideally normalized using a common taxonomy represented
and exchanged using standardized formats and protocols. Research is
needed in these areas; NIST and others are already working on some of
these issues. How the data should look (form/format) is the easy part;
what data is most useful is much harder.
Question 5b. Can too much information actually be a problem or can
there never be too little information when it comes to cybersecurity
incidents?
Answer. Since a cybersecurity incident investigation often starts
as an attempt to discover the true scope and scale of what transpired,
various data sources need to be synthesized. The issue is not
necessarily having more data, but the right data. We frequently see
cases where information collected and shared is useless. Without
context about the incident, it is difficult to abstractly predict what
might be needed in advance. There is inherent cost in extracting and
delivering the data. Hence, it is convenient to know what data is
available and to be able to request it on demand. Achieving this
enhanced situational awareness will require continued research and
pilot programs with data owners.
Question 6. How can legislation assist in facilitating capable,
scalable, and cost-effective cyber incident response for Government and
critical infrastructure?
Answer.
Encourage public/private cooperation and access to data for
empirical research.
Support training operators in the same context as they work.
Support scalable forensics capabilities.
Regularly recognize successes in cybersecurity and incident
response.
Successful response requires close cooperation between the
Government and the private sector, so as mentioned in question No. 1,
inclusion of the private sector in plans for incident response would
greatly improve response effectiveness. Expanding the scope of the
current policies to include plans for working with industry would allow
for more timely and capable responses. Cooperation should also include
access for innovators to incident data, which will result in better,
scientifically validated solutions. Additionally, the Government must
continue to engage the community at large to maintain perspective on
what currently exists, both in terms of technological gaps and
solutions.
People who respond to cyber incidents must be adequately trained.
The Government needs a training solution that is scalable and cost-
effective, such as CERT's Virtual Training Environment (VTE) and X-NET.
Traditional training and education models still employ brick and
mortar classrooms to provide infrequent instruction directed at
individual students. These models simply cannot keep up with the pace
of change or provide successful and cost-effective mechanisms for
organizations to gain and maintain the real-world experience needed to
operate effectively in cyberspace. Civilian employees cannot use
production agency networks for operational training and ranges or
laboratory environments can be costly to develop, operate, and
maintain.
In addition to training and practice limitations, agencies
currently do not have any reliable capability to assess the operational
mission readiness of their cyber workforce. The current unit-level
cyber assessment mechanisms rely on artificial paper-based simulations
and ``cyber-add-ons'' to intra- and interagency exercises. Neither
approach provides for reliable mission-readiness evaluation and
reporting of workforce effectiveness.
CERT's VTE provides rich media instruction and hands-on training
labs to remote students over the internet. It enables students to
access high-quality training on security, computer forensics, and
incident response anywhere in the world, with only a web browser and an
internet connection. What's more, VTE is a cost-effective way to train
the workforce,\1\ and has no expiration date, allowing students access
to all training modules as often as they want and for as long as they
want after completing training. Students can continually return to the
module to practice and test the network, closing the gap between
learning a concept and using that concept.
---------------------------------------------------------------------------
\1\ High-Fidelity e-Learning: SEI's Virtual Training Environment
(VTE): TECHNICAL REPORT CMU/SEI-2009-TR-005 ESC-TR-2009-005: VTE was
used to deliver 38,157 hours of training for DISA during the period
from January 1, 2007 through October 31, 2007. The American Society of
Training and Development (ASTD) reports that the average cost per
learning hour delivered by its members in 2006 was $54.25. According to
the ASTD data, the value of VTE-delivered training is therefore
$2,070,017 ($54.25 per hour 38,157 hours = $2,070,017.25). The total
cost to DISA for the VTE-delivered training was $858,250. This
represents a cost savings to the DISA of $1,211,767 as compared to what
they could have expected to pay at prevailing industry average costs.
The total return on investment for the DISA is 141 percent.
(($2,070,017 - $858,250) / $858,250 = 141%).
---------------------------------------------------------------------------
CERT's Exercise Network (XNET) provides real-world experience
building and readiness evaluation via synchronous, team-based,
scenario-driven cyber exercises. Experience through routine practice is
known to be the decisive factor in how effectively individuals and
organizations respond during incidents and emergency situations. XNET
is designed to make this routine practice web-accessible for globally
distributed teams and units.
The Federal Government needs to address its current backlog of
cyber forensics data, as well as, collect forensics data in on-going
cases in a timely and cost-effective manner. To help augment the cyber
forensic capabilities of law enforcement the CERT program created the
Clustered-Computing Analysis Platform (C-CAP). C-CAP is designed to
support 200 concurrent computer examinations looking at 200 terabytes
of data, allowing for a massive, coordinated effort. Absent
catastrophic events, the C-CAP environment can offer underequipped or
overwhelmed agencies real-time additional resources. C-CAP is a state-
of-the-art forensics analysis environment that provides a complete
suite of tools for host-based and network investigations. C-CAP
augments scarce resources by allowing multiple users to view the same
data, either remotely or locally; while maximizing the application of
specialized computing resources to the forensic and incident response
missions. Analysts and investigators enjoy flexible, secure access to
high-performance systems, increasing productivity and facilitating
distributed collaboration. Designed specifically for forensics and
incident response analysis, this unique integration and packaging of
tools, accelerates the analysis processes, maximizes performance and
reduces costs. C-CAP is a flexible solution, allowing agencies to add
or remove components that are relevant to their particular needs. Its
unique centralized management interface allows organizations to rapidly
allocate platform resources to tasks or analysts. Scalable and cost-
effective, C-CAP can be customized to suit any organization, regardless
of size and mission.
Finally, we recommend that the Government recognize and reward good
examples of secure systems and practices. In the end, infrastructure
components need to be built more securely in the first place and by
highlighting those organizations who are doing it right, the Government
can incentive others. The Baldrige Program is administered by the
National Institute of Standards (NIST) and educates organizations in
performance excellence management and administers the Malcolm Baldrige
National Quality Award. This public-private partnership is helping
organizations achieve best-in-class levels of performance; identifying
and recognizing role-model organizations; identifying and sharing best
management practices, principles, and strategies. A similar program or
award in the area of security and resiliency could yield substantial
benefits.
Questions From Chairman Daniel E. Lungren for Leigh Williams
Question 1a. You describe a large number of items members of the
financial services sector undertake with respect to cybersecurity.
Can you compare these activities with those of the other sectors?
Answer. We are not in a position to compare the quality or quantity
of cybersecurity efforts in other sectors to financial services, but we
can identify some similarities and differences. As a similarity, we
recognize that individual companies in telecommunications and
information technology invest heavily in cybersecurity and resiliency.
We understand that one difference is that financial institutions may do
more collaborative work because they are so technically and
commercially interconnected and because regulations tend to promote
standardization.
Question 1b. Which of these activities are the product of voluntary
action by the BITS community and which are the result of Federal or
State regulations?
Answer. At the institution level, most BITS members' cybersecurity
programs are primarily motivated by business and customer interests.
Regulations sometimes reinforce these motivations, but also sometimes
require slightly different solutions. For example, under Gramm-Leach-
Bliley, banks are required to have security programs that incorporate
specific elements and that are reviewed by their boards. Without the
regulation, the vast majority of banks would still have plans, but
perhaps with different mixes of elements, and with review processes
specific to their governance strategies. At the industry level--in
efforts such as the mobile, cloud, social networking, and malware
efforts mentioned in our June 24 testimony--virtually all of the
collaboration is purely voluntary.
Question 1c. What is the cost of complying with these activities?
Answer. We do not have a specific estimate of regulatory compliance
costs in cybersecurity. We do believe, however, that elevated
compliance costs can crowd out risk management spending and investments
in innovation, and can increase costs to customers and reduce
institutions' returns.
Question 2. Under the administration's proposal what new
cybersecurity activities would BITS members undertake that they are not
now doing?
Answer. Under the administration proposal, there would be at least
two ways in which BITS members could more effectively share information
with other sectors. First, because other sectors could be prompted to
produce more information and DHS would be tasked with aggregating it,
there would be more information available to exchange with our
colleagues in other sectors. Second, the safe harbor and
confidentiality provisions would reduce the risk of actively sharing
information with the other critical infrastructures and with DHS.
Question 3. You are endorsing the administration's legislative
proposal, which does not carve out the financial sector from its reach.
With this endorsement is it to safe to assume that the financial
industry will not be lobbying for a carve-out or any special treatment
if the administration's proposal moves forward?
Answer. BITS does not intend to advocate for the financial services
sector to be carved out. BITS and its members do believe that the
existing financial regulatory frameworks and the proposed approach will
have to be reconciled. As we testified, this could be accomplished, for
example, by recognizing where substantially similar requirements
already exist, by leaving substantial authority within the sector, by
requiring DHS to work through the sector-specific agencies and primary
regulators, or by DHS delegating authority back to the sector-specific
agencies and primary regulators.
Question 4a. Your testimony praises the administration's
legislative proposal for a variety of things like coordinating with
companies and other agencies; however, it was my understanding that
most, if not all, these activities are currently going on without this
legislation.
Which specific provisions of the administration's proposal will
cause BITS members to make security improvements beyond their current
activities and why is legislation required to get the BITS membership
to undertake these activities?
Answer. Yes, BITS members are already satisfying many of the
requirements of the administration's proposal. The value of the
proposal does not arise primarily from BITS members individually
improving their security programs. Much of the value arises from
companies in multiple industries and Federal agencies with various
missions working in closer cooperation on common problems. We think
this is happening reasonably well within our sector, but we see room
for improvement between sectors.
Question 4b. How much will these legislatively-mandated activities
by BITS members improve security?
Answer. While the mandates in the proposal may improve BITS
members' cybersecurity practices, we see much of the potential
improvement coming from enabling more voluntary collaboration. For
example, as noted above, we would anticipate improved information
sharing and consequently better collective security among multiple
sectors, including financial services.
In closing, we reaffirm our commitment to addressing this critical
issue, and thank the committee for its active engagement. Please feel
free to contact me with any further questions or concerns.
Questions From Chairman Daniel E. Lungren for Larry Clinton
Question 1. Playing Devil's advocate, if critical infrastructure
must be regulated, what do you think that regulations should look like?
What is an appropriate framework for regulations?
Answer. Although the ISA generally supports market incentives as
opposed to Government regulation as the best way to spur the needed
investment in cybersecurity, this is not an absolute.
In fact ISA has always advocated a multi-tiered system with
appropriate regulation mixed with market incentives. This approach is
developed more fully in the ``The Cyber Security Social Contract:
Policy Recommendations for the Obama Administration and the 11th
Congress'' (2008) and the ``Social Contract 2.0: A 21st Century Program
for Effective Cyber Security'' (2009)--both attached.
The key consideration is that cybersecurity is not simply an ``IT''
issue but an enterprise-wide risk management issue. If we are
considering cybersecurity as a risk management issue we need to assess
not only the technical considerations, but also the economic
considerations. Research has consistently demonstrated that cost is the
single biggest barrier to implementing effective cybersecurity
standards, practices (see CSIS and Pricewaterhouse Coopers studies
cited in my written testimony) and technologies which other research
has demonstrated to work (see NSA testimony, PWC survey, and Verizon
///Secret Service studies cited in my written testimony).
Where regulation is an inherent part of the economics of an
industry, such as in many critical infrastructures (electricity, water,
nuclear power etc. as well as some element of the financial system)
than the traditional regulatory structures may be an effective tool for
promoting appropriate investment in cybersecurity. Indeed in some
industry sectors of great interest to cybersecurity policy makers
regulation could be a more effective mechanism than a market incentive
if, as in the case of water systems for example, there really is no
market.
Of course many of these entities are regulated at the State and
local, not Federal level. Moreover, as the decision making devolves to
lower levels of Government more localized issues may evolve. For
example a State PUC may be resistant to approving investments by a
power company for fear of the effect this may have on local utility
rates which could have political complications for members of the State
commission. However the Federal Government has long history in finding
ways to provide incentives to the States and localities to adopt
policies in the National interest.
However even in some of these regulated sectors, market incentives
may still be a better mechanism than regulation. The regulatory
structure in most instances is too slow to keep up with the pace of
cyber attack vectors which change with the speedy evolution of
technology. Also regulation tends to push entities to achieve minimal
compliance whereas we may need a more aggressive effort on the part of
enterprises not just to comply with minimum standards but to
affirmatively look for malware and cooperate with broad industry
sectors, and possibly beyond in information-sharing activities (see
paper on information sharing by Jeff Brown in the attached Social
Contract 2.0).
For many of these sectors a more effective mechanism may well be
the use of streamlined regulation wherein outdated provisions or
redundant audit requirements could be offered in return for investment
in more aggressive methods of cybersecurity including intensive
internal monitoring of unauthorized outbound traffic and participation
in creative and more modern models of information sharing than are
currently being operated by DHS (see Brown paper cited above).
Question 2. We have had a public-private partnership for several
years yet the cyber problem continues to grown, doesn't that indicate
that the model doesn't work?
Answer. To begin with I'd suggest this is a non-sequitur. The
reason the cyber problem has grown is not that the partnership has
failed but because the current incentive structure massively favors the
attackers. Cyber attacks are cheap, easy to acquire, and can generate
massive profits. While cyber defense is a generation behind the
attackers, it's difficult to justify ROI since metrics for prevented
attacks are impossible to generate and cyber criminals are rarely
caught.
Moreover, both the Cyber Space Policy Review and the most recent
Verizon/Secret Service study have demonstrated that the market has
already produced adequate mechanisms to prevent or stop most attacks
which suggests the market is working (indeed most attacks are currently
stopped--just too many still get through).
That said, the ISA has said from the first publication of the
National Strategy to Secure Cyber Space (2002) that the missing link in
the public-private partnership is the lack of incentives. The public-
private partnership is the right model but it needs to be evolved to
meet the modern threats and more fully implemented--especially by the
Government partners.
Research cited above as well as in my written testimony has long
demonstrated that that only a substantial minority (probably between
30% and 40%) of enterprises have what may be called a natural ROI for
security investment. When such as natural confluence occurs then
private sector entities will make adequate security investment.
However, as illustrated in the pan-association White Paper on
cybersecurity (cited in my written testimony and also attached) in most
instances the public sector and private sector assess risk differently.
In short form, for most of the private sector security is simply an
economic consideration. If you own a warehouse and 10% of your
inventory is ``walking out the back door'' every month, you will not
buy the cameras, hire the guards, etc. to solve your security problem
if your study shows that it costs 11% to do so. That is a good risk
management decision from a private-sector perspective.
The public sector has economic considerations, but also additional
non-economic considerations (National security, privacy, politics etc.)
and thus may have a lower-risk tolerance than their private partners
because they simply assess risk differently.
However, as the trade associations who signed onto that paper have
attested, we recognize that in an interconnected cyber world the
private sector may be required to take on new, non-economic, and
traditional public sector responsibilities with respect to
cybersecurity.
Therefore the public-private partnership which has heretofore
ignored the economic aspects of cybersecurity needs to evolve into a
fuller and more sustainable model which includes Government finding
ways to offset the non-economic investments it would like private
industry to make in the interests of broad National security.
Additionally, the fact is that the public sector has not been
faithful to following through on their responsibilities in the
partnership as laid out both in the NIPP and the Cyber Space Policy
Review. For example, markets cannot function without information--a
central tenant of Wall Street--but it is well-acknowledged that despite
millions spent on supposed Government information-sharing programs most
such shared information is of little or no use to the private sector.
Government still does not share the actionable threat information that
would allow among other things for a proper assessment of cyber risk
and assist greatly in making the proper investments.
Industry is not blameless here also. As illustrated in two
additional volumes attached (``50 Questions Every CFO Should Ask About
Cyber Security'' and ``The Financial Management of Cyber Risk'')
industry, largely due to antiquated corporate structures and
misunderstandings about the true nature of the cyber threat tends to
misunderstand the true financial implications they are dealing with.
These and other issues explain why the partnership has not fully
worked are more extensively detailed in the pan-association white
paper.
Question 3. Mr. Clinton, you advocate for the providing of market
incentives to the private sector to improve cybersecurity, given the
significant budget issues the Congress faces how can we afford to
provide market incentives for cybersecurity to the private sector?
Answer. One of the most persistent problems with digital economics
is that everyone wants to capture the profits of digital technology but
resists reinvesting a small portion of these profits in securing the
technology that is generating them.
Nearly every company in the world has by now factored into its
business plan the wonders of digitalization--web-based marketing,
international supply chains, VOIP instead of traditional
telecommunications, and remote workers. Yet, as described above we are
not getting the investment in cybersecurity that we should.
This is true for the Federal Government as well. For example the
Obama administration has announced a ``cloud first'' strategy for the
Federal electronic systems that they claim will save them between $20-
50 billion a year. Some of that money ought to be being plowed back
into system-wide--not just Government--cybersecurity.
However, assuming that none of this money will be invested in
market incentives there are still many levers the Federal Government
can use to generate more private cyber investment which require little
or no Government spending. Ironically, many of these incentive
structures are widely used in other areas of our economy; we simply
have not yet applied them to cybersecurity.
The key is to reduce Government-induced costs on industry, rather
than provide direct Government subsidies such as with tax incentives.
For example many companies may be attracted to making greater
cybersecurity investments in return for lower liability. Less stringent
liability costs the Government nothing but cold be perceived as an
economic benefit to industry.
Another example is streamlined regulations, or as appropriate
accelerated permitting and approvals. For example many enterprises are
buckling under redundant cybersecurity auditing requirements. If the
Government could develop a sound baseline audit to simply remove the
redundancy this could be offered as a carrot to enterprises that
demonstrate investment in proven effective e-cybersecurity techniques
such as those identified in the Verizon/Secret Service study cited in
my testimony.
On a broader scale there are numerous outdated analogue-based laws
(see Cyber Space Policy Review Appendix A) which could be modified
possible with reduced cost to industry.
Government procurement--not just for IT equipment--could also be
tied to more stringent cybersecurity on the part of firms that compete
for Government contracts, or access existing (not additional)
Government spending programs (e.g. small business loans--and all the
TARP money should have come with cybersecurity requirements). In these
cases we are not talking about Government spending more, we are simply
talking about who gets the spending the Government is making--weigh it
more heavily in terms of the compelling National interest of
cybersecurity. No new spending required.
There is also a great deal that can be done to stimulate the cyber
insurance market. With a broader insurance market we can off-load much
current Government risk to the private sector. Moreover, insurance
(discounts) are a major motivator of all sorts of pro-social behavior
from smoking reduction to improved driving and building safety. ISA has
done a fair amount of work on how to use insurance better ranging from
some relatively immediate items such as sharing information leading to
lower rates and greater uptake (due to more realistic risk assessments
and pricing) to broader programs dealing with National re-insurance.
The Social Contract documents (attached) provide some additional
examples.
Question 4. If as you say we know how to prevent or mitigate most
basic cyber attacks by use of current standards and activities why
don't we just mandate that companies do these best practices?
Answer. We can't just put seatbelts on the internet and think we
have solved the problem.
As identified in answer No. 2, the problem is that there are
massive incentives right now favoring the attackers.
Yes we have come up with ways to deal with most current attacks,
but the attack methods will continually evolve.
ISA is not interested in solutions; it is interested in creating a
sustainable system of cybersecurity.
To do this we need a much more dynamic motivator than Government
regulations, we need to use the market.
As described in greater detail in Chapter 1 of each of the Cyber
Social Contract documents attached, the Government regulatory model
invented to address the hot technology of 2 centuries ago--the
railroads--is not going to work for the 21st Century problem of
cybersecurity. We need a more active model which will keep up with
attacks, can be applied internationally, will not provide a roadmap to
the attackers and generate an atmosphere of foe compliance (equivalent
to campaign finance laws which everyone complies with and no one thinks
actually addresses the ``problem'' they are supposed to solve).
Regulations, (outside of those sectors for which the regulations
are part of the inherent economy of the sector as described in answer
1) will be too slow, outdated quickly, and too minimalistic to address
the modern problem we face.
Question 5. If companies are losing so much money due to cyber
attacks, why are there not already enough incentives for them to invest
to stop the attacks?
Answer. Part of this answer was addressed in the answer to question
3, above, where we discussed the fact that industry and the Government
assess risk in fundamentally different ways with industry, concerned
almost entirely about the economics of the situation, have a greater
risk tolerance than the public sector.
However there are many other problems. For example, it is very hard
to make truly accurate assessments of economic cyber losses for a
variety of reasons including the fact that for sophisticated cyber
attacks one may not know they have been the victim of the attack until
long after it has occurred because as in the case of the loss of
corporate IP, (the largest economic loss) the property is not stolen in
the physical since--it remains--it's just a copy has been made and
maybe being used to create a clone product or service.
In still another complication we have the ``interconnection
problem''. Due to the inherent interconnectedness of the internet it is
possible for a thief to steal your data that happens to be residing on
my system (I may not even have a direct relationship with you--I could
be a sub-contractor to a subcontractor--with little or no incentive to
protect your data which is valuable whereas my own data may not be as
valuable so I don't invest in security adequate to your needs.
Additionally, we have the problem with poor appreciation of actual
financial risk as described above.
Finally, there is the fact that in the current economic climate
business are being forced to make themselves ever more efficient
including cutting costs by adopting less secure technologies. VOIP,
international supply chains and cloud computing are all examples of
technologies that are increasing our cyber risks but are being widely
deployed (including by the U.S. Federal Government) despite their
security flaws due to the irresistible economic imperatives we all
face.
Government's job ought not to be to punish the victims of cyber
attacks who are forced to compete in the digital world we now inhabit
but to use the mechanisms at its disposal creatively, as described
above to assist enterprises in securing our Nation's system in a
sustainable and economically sensible way.
|
NEWSLETTER
|
|
Join the GlobalSecurity.org mailing list
|
|
|
