Spafford HASC Cybersecuirty TestimonyDownload PDF
"Cyber Security, Information Assurance and Information Superiority"
27 October 2005
Eugene H. Spafford
Professor and Executive Director
Purdue University Center For Education and Research in Information Assurance
and Security (CERIAS)
Chair of The U.S. Public Policy Committee
of The Association For Computing Machinery (USACM)
Member of the Board of Directors
of the Computing Research Association (CRA)
Table of Contents
Summary of Threats 2
Specific Threats 3
Exacerbating Conditions 4
Some Recommendations 7
Questions from the Committee 9
Thank you Chairman Hunter and Ranking Member Skelton for the opportunity to testify at this
hearing. Threats to information systems have been steadily growing in number and sophistication
over the last two decades. They currently present a substantial danger to the U.S. military, the
civilian government, industry, academia, and the general public. So many of these systems are
interconnected and dependent on each other that threats to one segment often spread to all the
others. Because many of these threats use victim computers to perpetuate the attack, it presents an
asymmetric threat to which U.S. computer systems are particularly vulnerable. In the remainder of
this document I will briefly outline a few selected aspects of the problems involved that are
especially important; a full treatment of all the issues would represent a major volume. The
complex interplay of the various components of our IT infrastructure mean that there are no simple
fixes for the problems we face. The best solution is to continually enhance and sustain our
capabilities to address cyber security problems, exactly as we do for other significant threats.
I will limit my comments to issues related to Computer and Network Defense (CND) issues, as the
specific questions posed to me were primarily about defense. Issues of CNA (attack) and CNE
(exploitation) are also involved in the overall context of Computer and Network Operations (CNO).
Details of CNA and CNE capabilities are generally highly classified, and I therefore am unable to
comment on them. However, given that potential foreign targets are using many of the same
products as those used by our own DoD, it would seem that other countries are currently as
vulnerable to cyber attacks as our military.
By way of self-introduction, I am a professor at Purdue University with a joint appointment in the
department of Computer Sciences and the school of Electrical and Computer Engineering. I also
have courtesy appointments in the departments of Philosophy, Communication, and the College of
Technology. I am also the Executive Director of the Center for Education and Research in
Information Assurance and Security. CERIAS is a campus-wide multidisciplinary institute, with a
mission to explore important issues related to protecting computing and information resources. We
conduct advanced research in several major thrust areas, we educate students at every level, and we
have an active community outreach program. CERIAS is the largest such center in the United
States, and we have a set of affiliate university programs working with us in a number of states,
including Illinois, Iowa, North Carolina, the District of Columbia, Ohio, Virginia, Idaho, and New
York. CERIAS also has a close working relationship with a dozen major commercial firms and
In addition to my role as an academic faculty member, I also serve on several boards of technical
advisors, including those of SignaCert, Unisys, Microsoft, DigitalDoors, and Open Channel
Software; and I have served as an advisor to Federal law enforcement and defense agencies,
including the FBI, the Air Force and the NSA. I was a member of the most recent incarnation of the
President’s Information Technology Advisory Committee (PITAC) from 2003 to earlier this year.
I have been working in information security for 25 years.
I began this document by listing my affiliations with ACM and CRA. This testimony is not an
official statement by either organization, but is consistent with their overall goals and aims. ACM is
a nonprofit educational and scientific computing society of about 80,000 computer scientists,
educators, and other computer professionals committed to the open interchange of information
concerning computing and related disciplines. USACM, of which I serve as the chair, acts as the
focal point for ACM's interaction with the U.S. Congress and government organizations. USACM
seeks to educate and assist policy-makers on legislative and regulatory matters of concern to the
computing community. The Computing Research Association is an association of more than 180
North American academic departments of computer science and computer engineering, industry and
issues that affect the conduct of computing research in the USA.
Summary of Threats
I will not attempt to summarize the magnitude of cyber threats currently facing our information
infrastructure, including our Department of Defense. There are many reports describing these
threats, including reports from the PITAC, the GAO, the National Academies, the Department of
Justice, and many commercial entities. From these reports the following general trends may be
• The number of reported attacks of various kinds is generally increasing annually;
• Attacks are becoming more sophisticated and more efficient;
• Few perpetrators are ever caught and prosecuted;
• An unknown (but probably large) number of attacks, frauds and violations are not detected with
• A large number of detected attacks are not reported to appropriate authorities;
• The problem is international in scope, both in origin of attacks and in location of victims;
• The majority of the attacks are enabled by faulty software, poor configuration, and operator
The Department of Defense operates many computers and networks that are little different from
those of other government agencies and private organizations. They are prone to some of the same
faults, misconfigurations, and operator mistakes. However, the DoD is an especially sensitive target
because of its role in securing the national defense and projecting US policy. As such, it is subject
to probing and attack by actors, such as national intelligence services, who might not be interested in
most commercial systems. However, vandals and anarchists will be as interested in attacking DoD
systems as they would in attacking other government systems. Criminal elements may well be
interested in attacking Defense systems to obtain sensitive information for sale or use, to steal
supplies (including weapons), to obtain intelligence that may impact their operations (e.g., military
involvement in drug smuggling interdiction), and to gather information for fraud (e.g., identity
To date, DoD systems have continued to fall prey to various forms of attack by these actors, and by
actors unknown. Computer viruses and worms have spread throughout Defense systems,
including some classified systems. Intruders have gained access to sensitive systems and data.
Information has been taken from DoD systems and used for purposes of fraud as well as common
espionage. Systems have been taken down and contaminated by unknown parties of foreign origin.
Network floods have disabled access to DoD systems. Undoubtedly the magnitude of the problem
is greater than has been openly reported, and more has occurred than has been detected. I have
every confidence that such incidents will continue to occur.
If this variety and magnitude of incidents were to occur to physical resources – unauthorized
access to sensitive data, disclosure of personnel information, significant fraud, espionage, and
degradation of function – there would be widespread outcry, investigation, and significant
disciplinary action against those in charge. Unfortunately, we have developed an attitude and
culture that views failures and compromises of important computing systems as inevitable and
acceptable. This is dangerous, and threatens the future economic and military safety of the
There are many threats to our IT systems, and to DoD systems in particular. As the DoD employs
large amounts of COTS (commercial off-the-shelf) products, or uses partially modified COTS
products in GOTS (government off-the-shelf) products, they are prey to some of the same
vulnerabilities as we see in the private sector and elsewhere in government. The threats are largely
the same, although the specific targets and consequences may well be different.
Not only are Defense systems based on common COTS products and protocols, but they are often
interconnected with (and dependent on) the private sector. The vast majority of DoD network
communications flows over private telecommunications networks, for instance. Military logistics
systems are often directly connected to civilian suppliers so as to provide direct B-to-B (businessto-
business) ordering to enhance speed and reduce cost. Military analysts are connected to online
news and search sites for open-source intelligence. Email is interconnected. Other connections
and dependencies also exist: some known, and many largely unknown. It is important to realize
that the DoD, despite some efforts to the contrary, is intimately connected to the rest of the national
(and thus, international) IT infrastructure. Even plans for the future Global Information Grid (GIG)
will not serve to disconnect these systems completely.
There are thus many pathways for attacks to occur against DoD systems (as well as other
government systems), and such connectivity will continue to exist (and is actually necessary for
many reasons). In the following, I will outline, generally, where I have observed potential
weaknesses in DoD IT systems:
Malicious software such as viruses and worms. Software that is self-propagating uses the
connectivity and power of victim computers against themselves and others. The homogeneity
of DoD systems makes them especially vulnerable to well-crafted malicious software.1 A
carefully crafted computer worm could cause disruption of operation, denial of service, and
corruption of information on large numbers of DoD systems.
Insider threats. I have observed deficiencies in internal protection and counterintelligence
operations at many DoD and government facilities over the last decade. At the same time, the
military has experienced an increase in the use of civilian consultants and contractors,
increased collaboration and exchange with foreign partners, and an increase in US citizens
involved in acts of terrorism based on idealism. All of these increase our risk of insider
attacks against key systems and data. Without adequate internal safeguards against insider
threats – both technical and operational – we are needlessly exposing our systems and data to
damage and exploitation.
Infiltration. Our military and government rely on COTS products and contractors to equip
and staff our IT infrastructure. Consider that some of those products that are employed in
highly sensitive applications are being crafted, tested, packaged and supported by individuals
who would never be allowed into the locations where those applications are used because of
national origin, criminal history, and/or personal behavior. Furthermore, some of the
hardware and software components in use in critical applications are designed and produced
in countries that may be adversaries in future military or political conflict. These factors
enable “life cycle” attacks where key systems can be compromised during early
manufacture, shipping, and maintenance as well as end operation. We do not have the tools
or resources to thoroughly check these items to ensure that they do not have “hidden
features” or flaws that may be used against us. We need special attention and methods to
produce these supervisory systems and critical applications.
Denial of Service. We currently depend on our networks and computing infrastructure.
Effective denial of service attacks against key systems would be devastating. These attacks
could be totally IT-based, as in network flooding against DoD communications sites, to large
scale physical attacks such as EMP (electromagnetic pulse) weapons used against whole
theaters. Some systems in some sites are protected against such attacks, and there are some
alternative systems and networks in place. However, I question whether sufficient planning
and risk assessment have been uniformly performed on scenarios where significant loss of
capability might occur. Without capable alternatives, our military would be significantly
Data Contamination. Many of our DoD systems depend on large data stores. These include
textual data such as personnel records, geospatial data such as maps and targeting
coordinates, and image data including parts diagrams and surveillance photos. Too few of
these records are protected against subtle alteration by outsiders. This protection might occur
through the use of immutable storage media, internal consistency checks, and digital
signatures. Gross changes, such as deletion, could be quickly spotted and repaired through
the use of backups. There is a danger, however, of long-term, subtle changes that alter key
data without the changes being detected. For instance, alteration of a few targeting data for a
theater of operations might result in multiple civilian targets being attacked in error, leading to
political damage and loss of confidence in the military systems in use.
All of these threat classes are real, and examples of all but the last can be readily found in the open
literature. I have seen little reasonable planning to address these threats, either within the DoD or
within the US government in general. Instead, the approach that is in widespread use is to employ
greater control over patching of some known flaws, and increasing the strength of some perimeter
defenses. Neither of these approaches addresses the underlying problems, and neither
appropriately anticipates future threats.
It is difficult to anticipate new threats that will emerge. Many experts believe the increasing
adoption of new technologies without careful consideration of risks will open new avenues of
attack. Two examples are the increasing use of wireless networks and voice-over-IP (VoIP)
telephony. Each of these offers new opportunities for convenience, mobility, and cost savings.
However, both technologies are more easily disrupted and intercepted than the traditional
technologies they replace. Other technologies being considered for future use include sensor
networks, telepresence, and grid computing. Without careful consideration of risks and defenses,
these may provide new opportunities for enemies and criminals to attack our IT infrastructure.
Equally of note is the increasing sophistication of the average attacker. As more value becomes
accessible via computer networks, there is a greater criminal element involved in cyber activities.
We should not underestimate that criminal element, especially as it includes individuals from
countries around the world. Not only will they increasingly target government resources for their
own gain, but some of them will undoubtedly operate on a “for hire” basis. Operating from
countries with weak law enforcement and hostility to the United States, these individuals pose an
unconventional threat that the military is not well equipped to handle.
There are many factors that further enhance the vulnerability of DoD (and other) IT systems, and
that will continue to endanger us in the future. I believe these six are among the most significant.
Over-dependence on COTS products. Several decades ago, the US Department of Defense
was the world leader in the development of innovative software engineering methodologies for
producing safe, effective computing products. However, almost all of the funding and inhouse
research in these areas was abandoned because of concerns over cost, and the delay of
getting products fielded that met quality standards. There was also the belief that commercial
systems with more features were better than limited, special-purpose military systems. The
underlying assumption was that the marketplace would drive commercial firms into
developing better methods and better software.
Unfortunately, the market responded differently than anticipated. The vast majority of
customers continue to want fancy new features rather than high-quality, robust software.
Given the choice, the majority of consumers will not pay extra for enhanced security, nor will
they easily tolerate some of the limitations that such security would impose. The market has
moved to continue to satisfy those desires and biases. Thus, the military is presented little
choice but to acquire COTS products to satisfy IT needs, it is also forced to use products that
contain large numbers of flaws, and that are designed for a very different threat environment
than where they are used: most vendors design for supporting computer game playing at
home and not joint force command and control in a real battle! Furthermore, government
acquisition is usually influenced more by cost than by fitness for purpose. If branches of the
uniformed military services were to acquire weapons platforms in the same manner, the Army
would be strapping howitzers to pickup trucks, the Air Force would be dropping JDAMs
from two-seater propeller planes, and the Navy would be patrolling the oceans in converted
Furthermore the market effectively demands that vendors regularly introduce new features to
sell new product every few months or years. The new features add complexity, which adds
new flaws, and in turn, adds new methods of attack. The additional complexity also means
that the operators and maintainers of the systems have an ever more complex environment to
understand and protect.
Many of the IT systems in use in the DoD today are vastly more complex and feature-laden
than needed. The typical desktop computer has far more functionality than is needed for
most applications, and thus more vulnerability than is prudent. However, because the DoD
depends on market solutions and lowest cost bids, there is a vicious cycle which results in
continued dependence on the same COTS products that have caused so many problems to
What is needed are policies and tools that enable us to extract the most valuable aspects of
commercial systems, especially in low risk environments, so as to take advantage of
commercial innovation and economy of scale. At the same time, we need to understand when
and where to restrict the use of such systems so as not to increase our exposure to attacks.
No metrics. We have no good metrics to measure safety, security, and quality of IT products
in a general and meaningful way. Therefore, we have no good method of comparing systems
against each other, or of determining whether a system’s overall security posture is improved
by a configuration change. A simple count of the number of patches issued or applied is not
sufficient, as this usually only reflects the responsiveness of the vendor. Neither does a count
of the number of attacks detected and repulsed represent strength of defense, as this simply
reflects the detected attacks so far. We need metrics that measure the resistance to attack of
individual components, but more critically, we need metrics that can be used to evaluate whole
systems of components as it is systems that are attacked.
What few metrics we do have, such as the Common Criteria, are intended to be applied
against severely constrained configurations that are seldom actually deployed. Once options
are changed and configurations modified, the original evaluations are no longer appropriate,
and often mislead the uninformed about the security of their systems.
Lack of deterrence. Nationally and internationally we have almost no deterrence. Vandals
and criminals can and do attack our IT systems with impunity because they know there is
almost no chance of being caught unless they are exceedingly careless. We have primitive
forensic capabilities and insufficient resources devoted to investigation and prosecution. We
also have questions of jurisdiction domestically and internationally.
As a result of the lack of deterrence, acts of vandalism and cyber crime are on the rise. Not
only is this damaging in and of itself, it provides a screen for more malicious activities:
intrusion detection and analysis software is so overwhelmed with “chaff” that it may be
unable to pick out the slow, deliberate, and skillful acts of espionage and sabotage that may be
Lack of fallback alternatives. I have observed great reliance within the DoD and government
in general on IT. Too often there is no planning for how to proceed with critical mission
responsibilities with degraded or disabled IT resources. This may be a failure to adequately
envision and quantify risks, or where cost has driven decisions that amplify risk.
Under-investment in research. The current attitude within government as a whole, and in
DoD in particular, is that long-term research is an option that can be cut given other budget
needs. Simultaneously, the real and pressing needs of current patches and defenses get a
huge share of resources. As was stated in the PITAC report, Cyber Security: A Crisis of
Prioritization (and elsewhere), this shortening of the horizon means that we will be at a
disadvantage in years to come. Innovation cannot be scheduled, nor can it be “caught up”
with short deadlines and short-term increases in support. The research base needs to be
supported consistently, over time, to build a body of results and a community of scholars.
Cyber security (and many other areas of IT) does not have that community of scholars, nor
do those researchers have the funding necessary to innovate as needed. Insufficient funding
leads researchers to be more conservative, and less likely to address big problems. In the
long term, this means we will continue to expend massive resources on fixing badly-broken,
inappropriate systems rather than deploying more resilient, better defended systems.
There is an analogy here to what we have seen with Hurricane Katrina: it was known for some
time that an unusual event could be catastrophic, and that contingency measures needed to be
developed. However, full funding was not allocated as other needs appeared to be more
pressing. Unfortunately, when the crisis came, there was no way that the needed responses
could be put into place quickly enough to avert the full scope of the disaster. Similarly, the
more we put off investing in finding solutions to the cyber security problem in favor of short
term needs, the greater the damage will be when the disasters occur.
Consider that other countries are increasing their support for long-term IT research. This
may lead to a future where important patents and key capabilities are held in countries other
than the USA, and where we are forced to obtain critical resources for our defense from
potential adversaries because we do not have the expertise nor infrastructure domestically to
meet the need.
The PITAC report, and reports cited in it all speak to this issue. The PITAC found, in
particular, that a shortening of the horizon in research funded by DARPA and by other DoD
agencies coupled with changes in emphasis have had a negative effect on the field. This is
only expected to get more pronounced in the coming years unless definitive action is taken to
reverse the trend.
For example, figures compiled by the CRA show that DARPA funding for university-led
research has declined by $100 million since 2001 despite a nearly $1 billion increase in
DARPA’s budget over the same period. More relevantly to the subject of this hearing,
DARPA support for university-led computer science research has plummeted, from $198
million in FY 01 (adjusted) to $108 million in FY 04 (adjusted).
We should not depend on the marketplace to address this issue. US companies are driven
largely by near-term profits and results. As a result we have seen a decrease in long-term
intramural research by US industry, and a decrease in extramural funding of long-term
research. Additionally, as I mentioned above, non-US companies may become very
significant players in the marketplace, and we should be cautious about depending on their
These trends do not bode well for US defense.
Ill-informed application of classification. Over the last few years, there has been an
increasing trend by various governmental agencies to classify anything to do with cyber
security defense research, and to limit the participation of non-nationals in related research.
This is usually misguided and definitely counterproductive.
There is no question that research into technologies for cyber offense should be closely
guarded. However, given the extensive use of COTS products and commercial infrastructure,
the only way we will enhance DoD cyber defenses is if improvements are made in publicly available systems. It is not possible to do this while classifying the research and results!
This issue was also investigated and discussed at length in the PITAC report.
Related to this issue has been the growth of restrictions and obstacles to non-citizens
participating in research in cyber security and IT issues in general (e.g., greater difficulties in
obtaining visas, and proposed changes to the deemed export rules on technology). This is
counterproductive because many of the great advances made in the last few decades have
come from foreign grad students in our universities, and from non-citizen engineers at
companies and professors at universities who stayed in the US after getting their degrees.
Historically, we have attracted the best and brightest researchers in the world to come study
and pursue their academic and commercial dreams. If we continue to make the US an
unfriendly destination for those individuals, other countries will reap the benefits of their
inventiveness and intelligence.
The vast majority of vulnerabilities and risks to IT systems, whether within DoD or elsewhere,
are plainly visible to people working in the field. Restricting who can address these
vulnerabilities, or unduly classifying the results, will (in the usual case) only serve to limit our
ability to protect ourselves.
I encourage members of the committee to carefully read the PITAC Cyber Security report. I
participated in the research and writing of that report, and it goes into much more detail on the
problems and issues behind our cyber security deficit.
There are several actions that can be taken to reduce the threat to Department of Defense IT
systems. In the following two lists, I present some that I believe could have the most impact over
the longer term. The first list is of items that Congress can address directly:
1. Most importantly, increase the priority and funding for scientific research into issues of
security and protection of IT systems. This was the conclusion of the PITAC, and of
numerous other studies cited in the PITAC report. Too much money is being spent on
upgrading patches and not enough is being spent on fundamental research by qualified
personnel. There are too few researchers in the country who understand the issues of
information security, and too many of them are unable to find funding to support
fundamental research. This is the case at our military research labs, commercial labs, and at
our university research centers. Increased spending for research is an investment in national
defense and national economic competitiveness, and is not in the same category as many
other expenditures for basic and applied research.
2. Increasingly, decisions on acquisition and deployment are being made by procurement
officers rather than the individuals with better knowledge of the risks and needs for cyber
defenses. At a minimum, there needs to be an explicit and prominent role kept for the
designers and operators of systems to ensure that security needs are not trumped by arbitrary
3. Provide increased support to law enforcement for tools to track malware, and to support the
investigation and prosecution of those who write malicious software and attack systems.
This includes support for additional R&D for forensic tools and technologies.
4. Revisit laws, such as the DMCA (Digital Millennium Copyright Act), that criminalize
technology instead of behavior. It is extremely counterproductive in the long run to prohibit
the technologists and educators from building tools and studying threats when the “bad
guys” will not feel compelled to respect such prohibitions.
The following list is comprised of suggestions that should be considered carefully by those within
the military charged with ensuring cyber security and effectiveness of our forces.
1. Rethink the use of COTS products in mission-critical circumstances — the lowest cost is not
necessarily the most fit for use. At the least, investigate better methods of screening and
testing such products to ensure that they do not contain hidden, unwanted features.
2. Initiate significant research into the development of metrics for security and risk. Acquiring
systems based on cost as the primary criterion is not reasonable for mission-critical
applications. We need to be able to differentiate among different vendor solutions, and set
standards of performance. Understandable metrics are needed for components and for entire
systems of systems (although they are not sufficient on their own).
3. Emphasize the need for a systems-level view of information security. Assuring individual
components does little to assure overall implementation and use. This requires trained
personnel with an understanding of the “big picture” of IT security. Too often those who
design and specify the systems do not understand how they are actually used....or misused.
4. Establish research into methods of better, more affordable software engineering, and how to
build reliable systems from untrusted components. The military needs to reengage in this
domain to ensure that their unique and critical needs are met.
5. Explicitly seek to create heterogeneous environments so that common avenues of attack are
not present. This may require some extra expense at first, but eventually it may lead to
increased compliance with standards, increased innovation, and increased choice in the
marketplace, thus lowering costs while increasing security. If real standards (rather than de
facto standards) are developed and followed, interoperability should not be a concern.
6. Complementary to the previous recommendation is giving thought to different architectures
in appropriate circumstances that better meet policy objectives. For instance, rather than a
computer on each desktop, thin-client technologies based on a mid-size computer in a
centralized location may provide all the same mission-critical services, but remove many of
the dangerous aspects of distributed PCs. In this situation, patches need only be applied in
one location, and there is a greatly reduced possibility of untrained users loading untested
media or software.
7. Rethink the need to have all systems connected to a network. Standalone systems may not
receive all of the latest patches as soon as they come out. However, that alacrity may not be
needed as those systems can no longer be attacked over the network.
8. Reexamine the issues of the insider threat to mission critical systems – from obtaining
software produced by uncleared personnel offshore and in this country, from using COTS
products that are not designed for security and reliability, and from access and operation by
untrained or unsupervised personnel. The intelligence community already does this, but it
needs to be considered for wider use across the military.
9. Reexamine the automation of critical systems. Do we have adequate alternate methods of
processing if core systems become unavailable or inaccessible? Do we have adequate noncomputer
copies of critical data that can be used operationally and to verify the integrity of
online data? Have we automated a system that was flawed in the pre-automation form, and
we have simply carried over the flaws into the new IT-based versions?
10. Establish better incentives for security. The current climate in many military commands and
government agencies is to penalize operators for flaws, thus leading many of them to dread
enhancement and exploration of better security.
Questions from the Committee
Given the category of threats, what is the worst case scenario to U.S. national security interests?
I am unsure what the worst case might be, but consider a time 15 years from now, where there is
considerable international tension with a large country in Asia. Because of the underinvestment in
long-term research in the US, we have been forced for the last several years to buy many of our
advanced computing systems for the military and for civilian infrastructure control from companies
located in that country. Outsourced maintenance is being done by companies in countries that are
within the sphere of influence of this major country.
The tension increases, and our adversary invades a neighboring country – one of our long term
allies. As that happens, large portions of our military command and control systems start crashing
mysteriously, our databases of targeting information become corrupted, and some major logistics
systems start issuing contradictory and incorrect orders for transportation and acquisition.
Meanwhile our civilian power grid goes down, as does part of our communications. Outages
randomly continue to occur for nearly a week including multiple failures in the SCADA systems.
Hundreds die across the US from accidents and incidental problems. Some investigation suggests
that these may be caused by hidden capabilities in the hardware and software of the controlling
systems, but we don’t have the tools or expertise to fully investigate the systems. The majority of
our bases domestically do not have sufficient power to operate at capacity, and in any event, the lack
of civilian power has taken out long-haul networks and satellite stations thus crippling our
communications as well. What units are able to respond find that key data for targeting, command
and control, and force protection are no longer correct or available, and no immediately usable
backup systems are available.
The President decided to respond with limited, precision military force to the invasion of our ally,
but the military is unable to mount a coordinated response because of the disruptions. By the time
some order is restored, our ally has fallen. We have lost an ally, other allies have lost faith in us,
and we have lost confidence in our own systems. Our economy has taken a major hit from the
national power and telecommunication outages and the public is both fearful and angry. We
receive an indirect message through third parties suggesting that if we don’t accept our adversary’s
new “province” then the US might experience even worse mysterious outages affecting power,
transportation, and finance. The military cannot protect us against these threats.
This case could be made even worse if it occurred simultaneously with a natural disaster or military
attack (or both!).
Given the category of threats, what are the most likely scenarios the U.S. might face?
In the near term, I expect we will continue to face more acts of vandalism, crime and espionage.
These will continue to cost a great deal of time and money to address. The severity and number of
these attacks will increase over time.
There are too many variables to project far into the future. However, some of the most effective
attacks involve clandestine exfiltration of information, and subtle alteration of internal data and
operations. Given the current state of the art and practice, we do not have strong assurances that we
have not already been victimized by such attacks, and that such attacks will not succeed in the
Are those scenarios preventable by nonmilitary means? If so, by what means?
I have provided a list of suggestions, based on my experience and research, in the preceding section.
Can those scenarios be addressed/mitigated by nonmilitary means? If so, by what means?
Again, this is addressed by the earlier section.
What other government departments and agencies (federal, state, and local) are involved in
addressing the scenarios? What roles do they play (lead, supporting) and what resources do they
Because of the interconnection and interdependency of our networks and computing systems, the
list of participating agencies and departments is quite large. Obvious lead agencies are DHS and
the FBI. Perhaps less obvious but very important are NIST and the NSF for the roles that they
play in supporting research and collaboration. The DOE, DOJ and NIH all have roles, as do the
police and homeland security departments of each state.
The key insight to answer this question is that the problem is one of a multi-pronged threat: attacks
by criminals, vandals, terrorists, anarchists, spies, and military agencies are all possible, and may not
be distinguishable before, during or after they occur. Additionally, those attacks may be specific or
indiscriminate, and they are likely to be committed against civilian as well as military targets.
To effectively defend the nation, we need to continually invest and promote R&D into both short
term and long term defenses for all sectors; we must obtain and deploy systems appropriate to their
mission, without undue functionality, and chosen for robustness rather than cost; we must
effectively investigate and prosecute misuse when it occurs to discourage other misuse and keep the
“noise” low; and we must continue to understand that all our systems are interconnected and
What kinds of military capabilities are useful for addressing the threat? (The panel should
identify two or three alternatives.)
I am unable to think of a military capability that will be adequate to deal with a decentralized
network-centric attack by a non-national entity.
Which of these capabilities does the United States currently possess? Which of these capabilities
is the United States currently developing? (Qualitative)
I do not have enough information to properly answer this question.
We are badly underinvested in long-term research for defenses, and we are badly undercapitalized
in the area of investigation and forensics.
What military capabilities does the United States possess in sufficient quantity to address the
threat? What military means does the United States have an excess of to address the threat?
I believe my answers above address this question.
What other question(s) should be asked?
I believe my other responses address this question.
1. I will conclude this testimony by reiterating the concluding statement made in my testimony before
a subcommittee of this committee on the 24th of July, 2003:
It is clear that we have deficiencies in our cyber defenses. Malicious and incorrect software
pose particular threats because of their asymmetric potential — small operators can exercise
large and devastating attacks on our defenses. The situation cannot be remedied simply by
continuing to spend more on newer models of the same systems and defenses that are
currently deficient. It will require vision and willingness to make hard choices to equip our
military with the defensible IT systems they deserve.
I will be happy to expand on any of these points, now or in the future
Thank you again for the opportunity to testify.
I received many suggestions from colleagues when composing this testimony. I wish to
acknowledge the people listed for their assistance. However, the content and opinions expressed
are my own, and the presence of these names should not be construed as endorsement of any of the
statements herein: Annie Antón, William Arbaugh, Randy Bond, Carla Brodley, Charles
Brownstein, Simson Garfinkel, Tim Grance, Peter Harsha, Peter Neumann, Ken Olthoff, David
Padgham, Dan Reed, and Robin Roberts.
1 I provided extensive written testimony on the threat of malicious software to DoD computers to a subcommittee of this committee on 24 July 2003.