The Cybersecurity Mess

The Cybersecurity Mess

Simson L. Garﬁnkel
Associate Professor, Naval Postgraduate School
April 25, 2013
“The views expression in this presentation are
those of the author and do not reflect the official
policy of the Department of Defense or the US
Government.”
1

NPS is the Navy’s Research University.

Monterey, CA — 1500 students
• US Military (All 5 services)
• US Civilian (Scholarship for Service & SMART)
• Foreign Military (30 countries)
• All students are fully funded
Schools:
• Business & Public Policy
• Engineering & Applied Sciences
• Operational & Information Sciences
• International Graduate Studies
NCR Initiative — Arlington, VA
• 8 offices on 5th floor, Virginia Tech building
• Current staffing: 4 professors, 2 lab managers,
2 programmers, 4 contractors
• OPEN SLOTS FOR .GOV PHDs!
2

“The Cybersecurity Risk”,
Communications of the ACM, June 2012, 55(6)

3

I have spent 25 years trying to secure computers...

APHY

ed. (Law & Business, Inc.,

, by Roger Tighe (Warren,

r Asset-Based Lending, by
ty, 1983)
Factoring and Finance, by
Co., New York City)

usiness Lawyer 1891 (1979)

iness Lawyer, by Harry J.
)

the scrutiny of counsel is
commitment letter and the
An Introduction to
Computer Security
1991 2000
nnection with the closing,
in advance and continue [Part 1]
; third, when any amend-
and, finally, the instant a Simson L. Garfinkel
point counsel should be
nd correspondence and, in "Spies," "vandals," and "crackers" are out there,
ve positions of competing
waiting to get into-or destroy-your databases.
THE LENDER'S HANDBOOK
hiladelphia, 1986)

AWYERS MUST UNDERSTAND is- Lawyers today must automatically
L the protection of their own inter-
for
sues of computer security, both recognize insecure computer systems
and lax operating procedures in the
ests and the interests of their clients. same way as lawyers now recognize

39

>

Sept. 1987

2006 2006
...and I have given up!
4

Today’s systems are less secure than those of the 1970s.

Reasons:
• Computers are more complex — more places to attack them.
• There are multiple ways around each defense.
• It’s easier to attack systems than defend them.
• It’s easier to break things than to ﬁx them.

5

We expect computers to crash...

… expect them to be hacked.
7

I start every day with...

[ISN]
Internet Security
News
8

[ISN] — infosecnews.org

9

[ISN] Anonymous-Backed Attacks Took Nasdaq Website
Offline
F
" rom: "InfoSec News <alerts@infosecnews.org>
S
" ubject: " [ISN] Anonymous-Backed Attacks Took Nasdaq Website Offline
"Date: "February 17, 2012 5:34:14 AM EST
" To: "sn@infosecnews.org
i
http://www.informationweek.com/news/security/attacks/232600975

By Mathew J. Schwartz
InformationWeek
February 16, 2012

The websites of the Nasdaq and BATS stock exchanges, together with the Chicago Board
Options Exchange (CBOE), were offline earlier this week after a hacktivist group with
apparent Anonymous ties targeted them with distributed denial of service (DDoS) attacks. But
while customers were intermittently unable to use some of the exchanges' websites, all said
that their trading systems weren't affected.

The attacks had been previewed the day before they were launched. In a post to Pastebin, a
group calling itself "the 'L0NGwave99' cyber group" said Sunday it was going to launch
"Operation Digital Tornado" in support of the "99% movement" Monday at 9 a.m. New York
time. A later message promised the same for Tuesday.

"The NASDAQ stock exchange besides a number of U.S. stock markets are going to face
http://infosecnews.org/ 10

[ISN] Most Small Healthcare Practices Hacked In The Past
12 Months
F
S
" ubject: " [ISN] Most Small Healthcare Practices Hacked In The Past 12 Months
i
http://www.darkreading.com/database-security/167901020/security/news/232601045/most-
small-healthcare-practices-hacked-in-the-past-12-months.html

By Kelly Jackson Higgins
Dark Reading
Feb 16, 2012

If you were wondering how safe your medical records are at your doctor's ofﬁce, then this
might make you sick: Ninety-one percent of small healthcare practices in North America say
they have suffered a data breach in the past 12 months.

The survey, conducted by the Ponemon Institute and commissioned by MegaPath, queried
more than 700 IT and administrative personnel in healthcare organizations of no more than
250 employees.

Among the ﬁndings: Only 31 percent say their management considers data security and
privacy a top priority, and 29 percent say their breaches resulted in medical identity theft.
"Cybercriminals are hunting for medical records," said Larry Ponemon, chairman and founder

[ISN] Air Force Special Operations Command eyes
Russian security software for iPads
F
S
" ubject: " [ISN] Air Force Special Operations Command eyes Russian security software for
iPads
i
http://www.nextgov.com/nextgov/ng_20120217_4350.php

By Bob Brewin
Nextgov
02/17/2012

When the Air Force Special Operations Command decided to buy 2,861 made-in-China Apple
iPad tablet computers in January to provide flight crews with electronic navigation charts and
technical manuals, it specified mission security software developed, maintained and updated
in Russia.

The command followed in the path of Alaska Airlines, which in May 2011 became the first
domestic carrier to drop paper charts and manuals in exchange for electronic flight bags.
Alaska chose the same software, GoodReader, developed by Moscow-based Good.iware, to
display charts in a PDF format on iPads. Delta Air Lines kicked off a test in August for
electronic flight bags and the carrier said it planned to use GoodReader software.


[ISN] 8 Lessons From Nortel's 10-Year Security Breach
F
S
" ubject: " [ISN] 8 Lessons From Nortel's 10-Year Security Breach
" Date: "February 20, 2012 3:16:51 AM EST
i
http://www.informationweek.com/news/security/attacks/232601092

By Mathew J. Schwartz
InformationWeek
February 17, 2012

It is every corporate security manager's worst nightmare.

News surfaced this week that Nortel's network was hacked in 2000, after which attackers
enjoyed access to the telecommunications and networking company's secrets for 10 years.

The intrusions reportedly began after attackers used passwords stolen from the company's
CEO, as well as six other senior executives, together with spyware. By 2004, a Nortel
employee did detect unusual download patterns associated with senior executives' accounts,
and changed related passwords. The security team also began watching for signs of
suspicious activity, but apparently stopped doing so after a few months. The full extent of the
breach wasn't discovered until 2010, by which time hackers had been accessing Nortel
secrets--from technical papers and business plans, to research reports and employees'

The cybersecurity mess: technical and social.

Most attention is focused on technical issues:
• Malware and anti-viruses
—Default allow vs. default deny
• Access Controls, Authentication, Encryption & Quantum Computing
• Supply chain issues
• Cyberspace as a globally connected “domain”

Non-technical issues are at the heart of the cybersecurity mess.
• Education & career paths
• Immigration
• Manufacturing policy

We will do better when we want to do better.

14

What do we know
think about
cybersecurity today?
15

Cybersecurity is expensive.

Global cybersecurity spending: $60 billion in 2011 pwc.com

• Cyber Security M&A, pwc, 2011 Cyber Security M&A
Decoding deals in the global
Cyber Security industry

172 Fortune 500 companies surveyed:
• Spending $5.3 billion per year on cybersecurity.
Cyber Security M&A
review

November 2011

• Stopping 69% of attacks.

If they raise spending...
Bloomberg Government Study »

• $10.2 billion stops 84% The Price of Cybersecurity:
Big Investments, Small
• $46.67 billion stops 95% Improvements

• “highest attainable level” » BY HELEN DOMENICI
Technology & Telecommunications
Policy Analyst

AFZAL BARI
Financial Analyst

Editor
ALLAN HOLMES
Analyst Team leader

95% is not good enough.
JANUARY 31, 2012

To contact the authors, e-mail:
hdomenici3@bloomberg.net and
abari1@bloomberg.net

© Copyright 2012, Bloomberg L.P. » www.bgov.com

16

Cybersecurity... is undeﬁned.

There is no good deﬁnition for “cybersecurity”
• Preventing computers from being “hacked”
• Using “network security” to secure desktops & servers
• Something having to do with cybernetics

There is no way to measure cybersecurity
• Which OS is more secure?
• Which computer is more secure?
• Is “open source” more secure?
• Does spending more money make a computer more secure? NO

17

Why the downward spiral?

Cybersecurity research does not make computers more secure
• “Reducing successful hacks” creates too big a target.
—Targets include data, apps, OS, network, human operators, hiring process,
supply chain, family members, ...
• Security research creates better attacks.

The environment is less secure:
• Increased interconnectedness
• Computers in more positions of trust
—Attacks today do more damage than attacks in the 1990s.

•

18

Cybersecurity is an “insider problem.”

bad actors
good people with bad instructions
remote access
malware

http://www.flickr.com/photos/shaneglobal/5115134303/

If we can stop insiders, we can secure cyberspace….
—But we can’t stop insiders.

19

Cybersecurity is a “network security” problem.

We can’t secure the hosts, so secure the network!
• Isolated networks for critical functions.
• Stand-alone hosts for most important functions.

http://www.flickr.com/photos/dungkal/2315647839/

But strong crypto limits visibility into network trafﬁc, and...

20

... stuxnet shows that there are no isolated hosts.

Every computer is connected to every other computer on the planet.
• USB sticks, DVDs, printers (“yellow dots”), scanners.
• Downloaded software (OS, applications), ﬁrmware, microcode
Every system is part of a computational ecology.

21

“Yellow Dots”
October 16, 2005

Secret Code in Color Printers Lets Government Track You
Tiny Dots Show Where and When You Made Your Print
San Francisco - A research team led by the Electronic Frontier Foundation (EFF) recently broke the code behind tiny tracking dots that
some color laser printers secretly hide in every document.

Sample closeup of Sample closeup of the
printer dots on a same dots showing only
normal printed page the blue channels to
make the dots more
http://seeingyellow.com/ visible.
22

Cybersecurity is a process problem.
“Security is a process,
Security encompasses all aspects not a product”
of an organization’s IT and HR
operations.

Microsoft Security Development Lifecycle

http://en.wikipedia.org/wiki/File:Bruce_Schneier_1.jpg

—Few organizations can afford SDL.
—Windows 7 is still hackable...
23

Cybersecurity is a money problem.

Security is a cost....
• ...Not an “enabler”
• No ROI

Chief Security Ofﬁcers are in a no-win situation:
• Security = passwords = frustration
• No reward for spending money to secure the infrastructure
• Money spent on security is “wasted” if there is no attack

24

Cybersecurity is a “wicked problem”

There is no clear deﬁnition of the wicked problem
—You don’t understand the problem until you have a solution.
There is no “stopping rule”
—The problem can never be solved.
Solutions are not right or wrong
—Beneﬁts to one player hurt another — Information security vs. Free speech
Solutions are “one-shot” — no learning by trial an error
—No two systems are the same. The game keeps changing.
Every wicked problem is a symptom of another problem

—Rittel and Webber, “Dilemmas in a General Theory of Planning,” 1973
—Dave Clement, “Cyber Security as a Wicked Problem,”
Chatham House, October 2011
http://www.chathamhouse.org/publications/twt/archive/view/178579

25

Why is
cybersecurity
so hard?
26

Cybersecurity has an active, malicious adversary.

The adversary...
—Turns your bugs into exploits
—Adapts to your defenses
—Waits until you make a mistake
—Attacks your employees when your systems are secure

27

Compiler bugs are security vulnerabilities!

The adversary chooses:
• What to exploit
• When to exploit it
• How to exploit it

We have seen:
• Optimizations can become
security vulnerabilities
• The same errors are
repeatedly made by different
programmers

What’s difference between
a bug an an attack?
—The programmer’s intent.

28

The supply chain creates numerous security vulnerabilities

App
Developers
3rd Party
Apps
Kits

Open
Source iOS

Apple
Developers
CPU

Wireless
Carrier

29

The attacker is smarter than you are…
… and has more time to find a good attack.

ACComplice: Location Inference using
Accelerometers on Smartphones
Jun Han, Emmanuel Owusu, Le T. Nguyen, Adrian Perrig, Joy Zhang Motion Trajectory
{junhan, eowusu, lenguyen, perrig, sky}@cmu.edu
Carnegie Mellon University
Mapped Points
GPS
Abstract—The security and privacy risks posed by smartphone Accelerometers are a particularly interesting case because of
sensors such as microphones and cameras have been well docu- their pervasiveness in a large assortment of personal electronic
mented. However, the importance of accelerometers have been devices including tablet PCs, MP3 players, and handheld gam-
largely ignored. We show that accelerometer readings can be
used to infer the trajectory and starting point of an individual ing devices. This array of devices provides a large network for
who is driving. This raises concerns for two main reasons. spyware to exploit.
First, unauthorized access to an individual’s location is a serious Furthermore, by correlating the accelerometer readings be-
invasion of privacy and security. Second, current smartphone tween multiple phones it is possible for an adversary to de-
operating systems allow any application to observe accelerometer

3 accelerometers
termine whether the phones are in close proximity. Because
readings without requiring special privileges. We demonstrate that
accelerometers can be used to locate a device owner to within a phones undergoing similar motions can be identified by their
200 meter radius of the true location. Our results are comparable accelerations, events such as earthquakes or even everyday
to the typical accuracy for handheld global positioning systems. activities like public transportation (e.g., bus, train, subway)

I. I NTRODUCTION
produce identifiable motion signatures that can be correlated
with other users. As a consequence, if one person grants GPS
access, or exposes their cellular or Wi-Fi base station, then they
no privacy
Location privacy has been a hot topic in recent news after it essentially expose the location of all nearby phones, assuming
was reported that Apple, Google, and Microsoft collect records the adversary has access to these devices.
of the location of customers using their mobile operating sys- a) Contributions: Our key insight is that accelerometers
tems [12]. In some cases, consumers are seeking compensation enable the identification of one’s location despite a highly
in civil suits against the companies [8]. Xu and Teo find noisy trajectory output. This is because the idiosyncrasies of
that, in general, mobile phone users express lower levels of roadways create globally unique constraints. Dead reckoning
concern about privacy if they control access to their personal can be used to track a user’s location long after location services
information. Additionally, users expect their smartphones to have been disabled [6]. But as we show, the accelerometer can
provide such a level of control [20]. be used to infer a location with no initial location information.
There are situations in which people may want to broadcast This is a very powerful side-channel that can be exploited even
their location. In fact, many social networking applications in- if location-based services on the device are disabled.
corporate location-sharing services, such as geo-tagging photos
b) Threat Model: We assume that the adversary can
and status updates, or checking in to a location with friends.
execute applications on the mobile device, without any special
However, in these instances, users can control when their
privileges except the capability to send information over the
location is shared and with whom. Furthermore, users express
network. The application will use some legitimate reason to
a need for an even richer set of location-privacy settings than
obtain access to network communication. This is easily accom-
those offered by current location-sharing applications [2]. User
plished by mimicking a popular application that many users
concerns over location-privacy are warranted. Websites like
download; e.g., a video game. In the case of a game, network
“Please Rob Me” underscore the potential dangers of exposing
access would be needed to upload high scores or to download
one’s location to malicious parties [5]. The study presented here
advertisements. We assume that the OS is not compromised,
demonstrates a clear violation of user control over sensitive
so that the malicious application simply executes as a standard
private information.
application. The application can communicate with an external
server to leak acceleration information. Based on the leaked
This research was supported by CyLab at Carnegie Mellon under grants
DAAD19-02-1-0389 and W911NF-09-1-0273, from the Army Research Office, information, the adversary can extract a mobile user’s trajectory
and by support from NSF under TRUST STC CCF-0424422, IGERT DGE- from the compromised device via data analysis.
0903659, and CNS-1050224, and by a Google research award. The views Our goal is to determine the location of an individual driving
and conclusions contained here are those of the authors and should not be
interpreted as necessarily representing the official policies or endorsements, in a vehicle based solely on motion sensor measurements. The
either express or implied, of ARO, CMU, Google, NSF or the U.S. Government
or any of its agencies. (a) Pittsburgh, PA
https://sparrow.ece.cmu.edu/group/pub/han_ACComplice_comsnets12.pdf
general approach that we take is to first derive an approximate
motion trajectory given acceleration measurements–which we
(b) Moun
978-1-4673-0298-2/12/$31.00 c 2012 IEEE Jun Han, Emmanuel Owusu, Thanh-Le Nguyen, Adrian Perrig, and Joy Zhang
discuss in §II. We then correlate that trajectory with map
"ACComplice: Location Inference using Accelerometers on Smartphones" In Proceedings
Fig. 7. Verification of map matching algorithm with the known startingSystems (a) and (b) show Experiment 2a
of the 4th International Conference on Communication point. and Networks
View, CA), respectively. The green (star) curve indicates the motion trajectory obtained from ProbIN. The red
(COMSNETS 2012), Bangalore, India, January 3-7, 2012.
blue (x-mark) curve indicates the ground truth (i.e., actual route traveled) obtained from GPS data.

30

Fortunately adversaries are not all powerful.

Adversaries are impacted by:
—Economic factors
—Attention span
—Other opportunities

You don’t have to run faster than the bear….

http://forums.corvetteforum.com/off-topic/2824193-i-only-have-to-outrun-you-polar-bear-tag.html 31

There are solutions to many cybersecurity problems...
... but we don’t use them.

30% of the computers on the Internet run Windows XP
• Yes, Windows 7 has vulnerabilities, but it’s better.

Apple users don’t use anti-virus.
• Yes, Apple tries to ﬁx bugs, but

Most “SSL” websites only use it for logging in.

DNSSEC
http://

Smart Cards

32

Many people liken cybersecurity to the ﬂu.

DHS calls for “cyber hygiene”
• install anti-virus
• update your OS
• back up key ﬁles

—“STOP, THINK, CONNECT”

33

A better disease model is obesity.

Making people fat is good business:
• Farm subsidies
• Restaurants
• Healthcare and medical utilization
• Weight loss plans
—Few make money when Americans stay
trim and healthy.

Lax security is also good business:
• Cheaper cost of deploying software
• Private information for marketing
• Selling anti-virus & security products
• Cleaning up incidents
—Few beneﬁt from secure computers

34

Nontechnical factors impacting cybersecurity.

Non-technical factors reﬂect deep divisions within our society.

• Shortened development cycles

• Education: General failure of our schools at science, engineering & math.

• HR: Inability to attract and retain the best workers.

• Immigration Policy: Foreign students; H1B Visa

• Manufacturing Policy: Building in your enemy’s factories is a bad idea.

Solving the cybersecurity mess requires solving these issues.

35

Short development cycles

Insufficient planning:
• Security not “baked in” to most products.
• Few or no security reviews
• Little Usable Security

Insufficient testing:
• Testing does not uncover security flaws
• No time to retest after fixing

Poor deployment:
• Little monitoring for security problems
• Difficult to fix current system when new
system is under development

36

Education is not supplying enough security engineers

Students are not pursuing CS in high school & college

Those going into CS are not pursuing security

Many of those studying CS are not staying in the country

37

73% of states require computer “skills” for graduation.
Only 37% require CS “concepts”

And teachers are poorly paid!
—Salaries for beginning & average teachers lag CS engineers by 30%
—Adjusting for cost-of-living and shorter work week.
• Linda Darling-Hammond, Stanford University, 2004
http://www.srnleads.org/data/pdfs/ldh_achievemen_gap_summit/inequality_TCR.pdf

38

High school students are not taking AP computer science!

Calculus

Biology

CS

http://www.acm.org/public-policy/AP%20Test%20Graph%202009.jpg

39

lment(by(Department(Type The number of CE degrees also increased
signiﬁcantly this year, among U.S. CE
CS departments who also give CE degrees. Degrees in the
Computer Science undergraduate enrollment is low...
signiﬁcantly among U.S. departments offering information degrees,
categorization of several institutions whose CS and I departments
2010-2011 CRA Taulbee Survey:
llment increased in aggregate among departments offering I
g those
t increased Figure 2. BS Production (All Departments)
ported both 30,000
CE and I
Number of Degrees

te, though 22,500
departments
e data 15,000

rees of all 7,500
be noted
are more 0
r of
95

97

99

01

03

05

07

09

11
of these
19

19

19

20

20

20

20

20

20
Source:(Table(3:(Bachelor’s(Degrees(Awarded(by(Department(Type

atile due to the small number of departments reporting. In
ecreased degree production, but Canadian response to the survey
d among Canadian departments reporting both years, there was an 40

Male 7,983 88.3% 1,856 88.2% 1,993 82.5% 11,832 87.3%

7% of Bachelor’s1,057 11.7% awarded to “nonresident alien”
Female degrees 248 11.8% 422 17.5% 1,727 12.7%
Total"Known"Gender 9,040 2,104 2,415 13,559
(12,800 to US citizens)
Gender"Unknown 246 0 1 247
Grand"Total 9,286 2,104 2,416 13,806

Table&5.&Bachelor’s&Degrees&Awarded&by&Ethnicity
CS CE I Total
Nonresident&Alien 524 7.0% 179 10.0% 78 3.6% 781 6.8%
Amer&Indian&or&Alaska& 39 0.5% 8 0.4% 16 0.7% 63 0.5%
NaBve
Asian 1,115 14.8% 337 18.8% 302 13.9% 1,754 15.3%
Black&or&AfricanKAmerican 274 3.6% 106 5.9% 151 6.9% 531 4.6%
NaBve&Hawaiian/Pac& 22 0.3% 7 0.4% 8 0.4% 37 0.3%
Islander
White 5026 66.9% 981 54.7% 1432 65.8% 7,439 64.8%
MulBracial,&not&Hispanic 104 1.4% 28 1.6% 3 0.1% 135 1.2%
Hispanic,&any&race 409 5.4% 146 8.1% 187 8.6% 742 6.5%
Total&Residency&&&&Ethnicity&
7,513 1,792 2,177 11,482
Known
Resident,&ethnicity&unknown 741 200 99 1,040
Residency&unknown 1032 112 140 1,284
Grand&Total 9,286 2,104 2,416 13,806

Table&6.&Total&Bachelor’s&Enrollment&by&Department&Type degrees.
—Most do not go on to advanced
& CS CE I Total
Avg.& Avg.&& Avg.& Avg.&&
41

50% of Master’s degrees awarded to nonresident aliens
(4960 to US citizens)
Table&9.&Master’s&Degrees&Awarded&by&Ethnicity
CS CE I Total
Nonresident&Alien 3,332 56.7% 776 72.6% 389 19.6% 4,497 50.4%
Amer&Indian&or&Alaska&
12 0.2% 0 0.0% 12 0.6% 24 0.3%
NaBve
Asian 753 12.8% 108 10.1% 245 12.3% 1,106 12.4%
Black&or&AfricanKAmerican 96 1.6% 13 1.2% 123 6.2% 232 2.6%
NaBve&Hawaiian/Pac&
19 0.3% 0 0.0% 6 0.3% 25 0.3%
Island
White 1533 26.1% 142 13.3% 1113 56.1% 2,788 31.2%
MulBracial,&not&Hispanic 8 0.1% 4 0.4% 4 0.2% 16 0.2%
Hispanic,&any&race 119 2.0% 26 2.4% 92 4.6% 237 2.7%
Total&Residency&&&&
5,872 1,069 1,984 8,925
Ethnicity&Known
Resident,&ethnicity&
320 88 205 613
unknown
Residency&unknown 419 26 17 462
Grand&Total 6,611 1,183 2,206 10,000

—We should let them stay in the country after they graduate
Table&10.&Total&Master’s&Enrollment&by&Department&Type
CS CE I Total
Department" 42

50% of PhDs awarded in 2011 to nonresident aliens
(642 to US citizens)
Table&13.&PhDs&Awarded&by&Ethnicity
CS CE I Total
Nonresident"Alien 634 48.1% 130 67.4% 44 37.0% 808 49.6%
Amer"Indian"or"Alaska"NaLve 2 0.2% 0 0.0% 2 1.7% 4 0.2%
Asian 171 13.0% 16 8.3% 14 11.8% 201 12.3%
Black"or"AfricanSAmerican 16 1.2% 1 0.5% 6 5.0% 23 1.4%
NaLve"Hawaiian/Pac"
4 0.3% 0 0.0% 0 0.0% 4 0.2%
Islander
White 465 35.3% 42 21.8% 52 43.7% 559 34.3%
MulLracial,"not"Hispanic 3 0.2% 0 0.0% 0 0.0% 3 0.2%
Hispanic,"any"race 22 1.7% 4 2.1% 1 0.8% 27 1.7%
Total"Residency"&""Ethnicity"
1,317 193 119 1,629
Known
Resident,"ethnicity"unknown 43 4 2 49
Residency"unknown 96 8 0 104
Grand"Total 1,456 205 121 1,782

—We did not train Russia’s weapon’s scientists at MIT during the Cold War.

43

Just 67/1275 (5%) PhDs went into Information Assurance
Computing Research Association

Table 14. Employment of New PhD Recipients By Specialty

Biomedica/ Other Science

Programming Languages/

Theory and Algorithms
Hardware/Architecture
Graphics/Visualization

Numerical Computing

Software Engineering
Computer-Supported

Information Retrieval

Information Systems
Artificial Intelligence

Information Science
Assurance/Security

Operating Systems

Social Computing/
High-Performance

Social Informatics
Cooperative Work

Human-Computer

Robotics/Vision
1828 L St. NW, Suite 800, Washington, DC. 20036

Informatics:
Databases /

Information
Computing
Interaction

Compilers

Scientific/
Networks

Other

Total
North American PhD Granting Depts.
Tenure-track 14 1 5 6 2 10 1 2 5 9 2 6 2 3 3 1 4 7 6 13 102 7.1%
Researcher 6 1 4 6 1 1 0 6 2 0 2 7 2 2 2 3 1 3 7 17 73 5.1%
Postdoc 38 1 12 17 4 12 0 20 7 5 2 12 7 7 14 6 3 10 30 34 241 16.8%
Teaching Faculty 2 1 1 0 0 1 0 1 1 2 1 1 1 1 0 0 3 4 4 4 28 2.0%
North American,
Other Academic
Other CS/CE/I
3" 0" 4" 1" 1" 1" 4" 2" 2" 0" 5" 6" 1" 0" 0" 0" 0" 3" 1" 18" 52" 3.6%"
Dept.
Non-CS/CE/I Dept.
http://cra.org

North American,
Non-Academic
Industry 64" 2" 49" 46" 41" 24" 20" 17" 40" 5" 6" 67" 29" 22" 25" 6" 12" 86" 32" 83" 676" 47.2%"
Government 7" 0" 5" 2" 6" 2" 5" 3" 8" 1" 2" 1" 0" 0" 2" 4" 1" 4" 2" 5" 60" 4.2%"
Self-Employed 0" 0" 0" 1" 0" 1" 0" 1" 0" 0" 2" 2" 2" 0" 1" 0" 0" 1" 1" 1" 13" 0.9%"
Unemployed 2" 0" 2" 1" 2" 2" 1" 0" 2" 0" 1" 3" 0" 0" 1" 0" 2" 0" 1" 3" 23" 1.6%"
Other 2" 0" 1" 0" 0" 0" 1" 1" 0" 0" 0" 1" 0" 0" 0" 0" 0" 0" 1" 0" 7" 0.5%"
Total Inside North
America
138" 6" 83" 80" 57" 54" 32" 53" 67" 22" 23" 106" 44" 35" 48" 20" 26" 118" 85" 178" 1,275" 89.0%"

Security should be taught to everyone, but we need specialists
44

Georgetown Prof: 50% of graduate students in sciences
are foreigners because salaries aren’t high enough.
Highest paying occupations:
• Medical: >$166,400
• CEOs: $165,080
• Dentists: $161,020
• Judges: $119,260
• …
• Computer Scientists: $115,070
• …
• Lawyers: $112,760
—Source: Bureau of Labor Stats

—Lindsay Lowell, Georgetown Institute for Study of International Migration.

45

Manufacturing policy

The USA did not build WW2 bombers in German aircraft factories.
46

Commercial Software radio Software radio Primary
programmer eavesdropper programmer risk
Determine whether patient has an ICD 4 4 4 Privacy

Security problems are bad for society as a whole...
Determine what kind of ICD patient has 4 4 4 Privacy
Determine ID (serial #) of ICD 4 4 4 Privacy
Obtain private telemetry data from ICD 4 4 4 Privacy
Obtain private information about patient history 4 4 4 Privacy
Determine identity (name, etc.) of patient 4 4 4 Privacy
Change device settings 4 4 Integrity
Change or disable therapies 4 4 Integrity
Deliver command shock 4 4 Integrity

… because computers are everywhere. TABLE I
R ESULTS OF EXPERIMENTAL ATTACKS . A CHECK MARK INDICATES A SUCCESSFUL IN VITRO ATTACK .

goal of understanding and addressing the potential security
risks of ICDs before future ICDs and other IMDs become
more complex and the potential security and privacy risks to
patients increase. However, we also firmly believe in disclosing
this information in an ethical manner that fully considers the
well-being of patients. We specifically and purposefully omit
details that would allow someone to use this article as a guide
for creating attacks against ICDs.
Paper organization. Section II gives a brief introduction
to ICDs, describes the security model we consider in this
work, and summarizes related work. Section III discusses
the process of intercepting and reverse-engineering an ICD’s
wireless communications, beginning with RF signal analysis
and culminating in readable plaintext. Section IV discusses
replay attacks that compromise device integrity by changing
stored information or therapy settings. Section V extends the

50 microprocessors discussion of zero-power defenses into the realm of device
2008: demonstrated wireless
design. Finally, Section VI offers concluding remarks. Fig. 1. Chest xray image of an implanted ICD (top right, near shoulder,

per average car
solid outline) and electrical leads connected to heart chambers (center of rib
II. BACKGROUND , M ODEL , R ELATED W ORK
attack on implantable pacemakers
AND cage, dotted outline).
This section summarizes the characteristics and medical
usage of a modern implantable cardioverter defibrillator. It also
introduces some of the equipment we used in our analyses. ways, including by inaction (failure to deliver treatment when
Following this introduction, we construct a security model that necessary) or by extraneous action such as a command shock
classifies potential adversaries in terms of their capabilities. when the heart is beating normally.
Finally, we summarize previous research that motivates and
Magnetic switch. Inside the ICD is a magnetic switch.
informs the methods and results of this work.
A magnetic field in proximity to this switch causes it to
close, which in turn causes the ICD to wirelessly transmit
• April 2012: McAfee announces successful wireless attack on insulin pump.
A. Implantable Cardioverter Defibrillators (ICDs)
An implantable cardioverter defibrillator (ICD) is a device
telemetry data, including electrocardiogram (EKG) readings.
(We discovered, however, that we can activate transmission of
that monitors and responds to heart activity. ICDs have modes
telemetry on our ICD solely with an RF command and without
—DDOS for the endocrine system!
for pacing, wherein the device periodically sends a small
electrical stimulus to the heart, and for defibrillation, wherein
the presence of a magnet; see Section IV.) In a clinical setting,
the magnetic field comes from a magnet in the programming
the device sends a larger shock to restore normal heart rhythm.
head, which is the component of the programmer that is placed
A physician surgically implants the ICD below the patient’s
in proximity to a patient’s implanted ICD. At the surface of
clavicle and close to the skin (Fig. 1). The physician also
one programming head we measured this magnet at 700 gauss.
implants electrical leads that connect the ICD to the heart
muscle. Post-surgery, a health care practitioner can use an Wireless communications. Our ICD wirelessly communicates
external programmer to perform diagnostics, read and write with the external programmer using the 175 kHz band, which
private data, and adjust therapy settings. A malfunctioning or is intended for short-range communications. Newer ICDs
maliciously configured ICD could harm a patient in multiple can communicate at both the 175 kHz frequency and in

This paper, copyright the IEEE, will appear in the proceedings of the 2008 IEEE Symposium on Security and Privacy. 3

47

[ISN] TV-based botnets? DoS attacks on your fridge?
More plausible than you think
F
S
" ubject: " [ISN] TV-based botnets? DoS attacks on your fridge? More plausible than you
think
"Date: "April 23, 2012 3:16:23 AM EDT
i
http://arstechnica.com/business/news/2012/04/tv-based-botnets-ddos-attacks-on-your-fridge-
more-plausible-than-you-think.ars

By Dan Goodin
ars technica
April 22, 2012

It's still premature to say you need firewall or antivirus protection for your television set, but a
duo of recently diagnosed firmware vulnerabilities in widely used TV models made by two
leading manufacturers suggests the notion isn't as far-fetched as many may think.

The most recent bug, found in a wide range of high-definition TVs from Samsung, was
disclosed on Thursday by Luigi Auriemma, an Italy-based researcher who regularly finds
security flaws in Microsoft Windows, video games, and even the industrial-strength systems
used to control dams, gas refineries, and other critical infrastructure. While poking around a
Samsung D6000 model belonging to his brother, he inadvertently discovered a way to
remotely send the TV into an endless restart mode that persists even after unplugging the

48

[ISN] ATM Attacks Exploit Lax Security
F
S
" ubject: " [ISN] ATM Attacks Exploit Lax Security
"Date: "April 23, 2012 3:15:54 AM EDT
i
http://www.bankinfosecurity.com/atm-attacks-exploit-lax-security-a-4689 http://krebsonsecurity.com/2011/12/pro-
grade-3d-printer-made-atm-skimmer/

By Tracy Kitten
Bank Info Security
April 19, 2012

Lax security makes non-banking sites prime targets for skimming attacks, like the ones that
hit eight hospitals in Toronto.

Earlier this week, Toronto police announced that eight area hospitals had been recent targets
for ATM skimming attacks. Over the past six months, authorities believe fraudsters targeted
these hospitals because of trafﬁc and the high-volume cash dispensers in these locations. But
security experts say the ATMs were more likely hit because they're easy targets.

"ATM placement in establishments like hospitals and 'cash only' enterprises seems to be an
afterthought to security, with the installation of ATMs in really remote areas of the building,
where fraudsters can easily tinker with skimming-device placement and retrieval without the
threat of immediate capture," says John Buzzard, who monitors card fraud for FICO's Card

49

Cell phones (&c) cannot be secured.

Cell phones have:
• Wireless networks, microphone, camera, & batteries
• Downloaded apps
• Bad crypto

Cell phones can be used for:
• Tracking individuals
• Wiretapping rooms
• Personal data

http://connectedvehicle.challenge.gov/
submissions/2706-no-driving-while-texting-
dwt-by-tomahawk-systems-llc

50

Five DARPA & NSF cybersecurity program managers walk
into a bar...
Major security breakthroughs since 1980:
• Public key cryptography (RSA with certiﬁcates to distribute public keys);
• Fast symmetric cryptography (AES)
• Fast public key cryptography (elliptic curves);
• Easy-to-use cryptography (SSL/TLS);
• Sandboxing (Java, C# and virtualization);
• Firewalls;
• BAN logic;
• fuzzing.

But none of these breakthroughs has been a “silver bullet.”
—“Why Cryptosystems Fail,” Ross Anderson,
1st Conference on Computer and Communications Security, 1993.
http://www.cl.cam.ac.uk/~rja14/Papers/wcf.pdf

51

There is no obvious way to secure cyberspace.

We trust computers…
—but we cannot make them trustworthy.
(A “trusted” system is a computer that can violate your security policy.)

We know a lot about building secure computers...
—but we do not use this information when building and deploying them.

We know about usable security…
—but we can’t make any progress on usernames and passwords

We should design with the assumption that computers will fail…
—but it is cheaper to design without redundancy or resiliency.
Despite the newfound attention to cybersecurity, our systems seem
to be growing more vulnerable every year.

52

The Cybersecurity Mess

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to The Cybersecurity Mess (20)

Recently uploaded (20)

The Cybersecurity Mess