The RISKS Digest
Volume 23 Issue 81

Monday, 28th March 2005

Forum on Risks to the Public in Computers and Related Systems

ACM Committee on Computers and Public Policy, Peter G. Neumann, moderator

Please try the URL privacy information feature enabled by clicking the flashlight icon above. This will reveal two icons after each link the body of the digest. The shield takes you to a breakdown of Terms of Service for the site - however only a small number of sites are covered at the moment. The flashlight take you to an analysis of the various trackers etc. that the linked site delivers. Please let the website maintainer know if you find this useful or not. As a RISKS reader, you will probably not be surprised by what is revealed…


Essex County NJ Jail locking-system failure
Charles Lamb
Cruise-control failures?
Robert Scheidt
TSA Finds Data On Air Passengers Lacked Protection
Amy Schatz via Richard M. Smith
RSA Finds More Flaws in RFID
Stephen D. Poe via Dave Farber
Sumitomo cyberattack
Tom Van Vleck
Clinical Healthcare IT, 'error', and safety
Richard Cook
Human error and computerized medical systems
Don Norman
Why IE is insecure: flawed logical thinking...
Craig DeForest
Re: Risks of long and short URLs
D.F. Manno
Risky US Bank Visa product
John Meissen
Important PITAC Cybersecurity report released
Gene Spafford
EEPI - Electronic Entertainment Policy Initiative
Lauren Weinstein

Essex County NJ Jail locking-system failure

<"Charles Lamb" <>>
Thu, 24 Mar 2005 01:17:08 -0500

According to an article in the 20 Mar 2005 issue of *The Star-Ledger*
(Newark NJ), New Jersey's Essex County Jail has experienced another failure
of its touchscreen based physical access control system.

Essex County Executive Joseph DiVincenzo sounds like he's never encountered
a competent systems engineer.  "These things happen. This is not the first
time it's happened. ...  It's happened a couple of times already, and it's
not going to be the last time either."  Police Benevolent Association Local
157 president Joe Amato has a more practical view, "Modern technology at its
finest. Who needs it? ... An old-fashioned turnkey operation would have just
been fine, but we spent millions for a high-tech computer-controlled jail
that isn't worth the contaminated dirt that it's built on."  One wonders how
well the system will function once the inmates get their hands on it.

Cruise-control failures?

<"Robert Scheidt" <>>
Wed, 23 Mar 2005 11:31:45 +0100

Recently in France a number of failures of "cruise control" systems
especially on recent models of Renault made cars have been reported, some
creating serious accidents (including a deadly one).  In general it is
reported that the car stays at his set speed and no matter what the driver
does, including cutting the ignition and breaking, the car continues at that

What's more surprising is that it is also reported that brakes become
ineffective (the brake pedal resists pressure).

I could imagine that the cruise control being probably under control of some
microprocessor, this microprocessor could "hang" due to some software
problem and therefore that everything it controls just stays as it is.
Especially in newer cars where fuel injection is completely electronically
controlled (no mechanical link between the gas pedal and the fuel injection

However, I have difficulties believing that the same microprocessor would
control the brakes and make them ineffective.  I wonder if somebody on this
board has some insight on how the electronic controls of modern cars are
designed and especially if a single component's failure (such as a common
microprocessor) could affect multiple functions (e.g., acceleration and

TSA Finds Data On Air Passengers Lacked Protection

<"Richard M. Smith" <>>
Fri, 25 Mar 2005 07:09:40 -0500

[Source: Amy Schatz <>, *Wall Street Journal*, 25 Mar
  2005, A4; PGN-ed],,SB111172077661889592,00.html?mod=todays_us_page_one

A new government report says officials in the Department of Homeland
Security didn't do enough to keep airline-passenger data secure when using
it to test a traveler-screening program.  DHS's Inspector General says the
Transportation Security Administration gathered 12 million passenger records
from February 2002 to June 2003 and used most of them to test the Computer
Assisted Passenger Prescreening System, or CAPPS 2, which was designed to
check passenger names against government watch lists. Passengers weren't
told their information was being used for testing.  TSA officials shelved
CAPPS 2 last year amid complaints it was an invasion of passenger
privacy. The agency has replaced it with a similar system, called Secure
Flight, which is being tested and is expected to debut in August.

The report raises concerns because Secure Flight ultimately will gather
private information, such as names, addresses, travel itineraries and
credit-card information, on anyone who takes a domestic flight. That effort
could be slowed by a Government Accountability Office study due Monday which
is expected to be critical of TSA's efforts to develop passenger-privacy

The report said TSA "did not ensure that privacy protections were in place
for all of the passenger data transfers" and noted that "early TSA and
[CAPPS 2] efforts were pursued in an environment of controlled chaos and
crisis mode after the Sept. 11 attacks."

Investigators also found TSA provided inaccurate information to the media
about the agency's use of real passenger records for CAPPS 2 testing and
wasn't "fully forthcoming" to the agency's own internal privacy officer
during an investigation into the matter. "Although we found no evidence of
deliberate deception, the evidence of faulty processes is substantial,"
investigators said.

RSA Finds More Flaws in RFID (via Dave Farber's IP)

<Stephen D. Poe <>>
Thu, 24 Mar 2005 10:56:00 -0500

By Jacqueline Emigh,, 23 Mar 2005

After uncovering a security weakness in a radio-frequency identification tag
from Texas Instruments Inc., researchers from RSA Security Inc.'s RSA
Laboratories and The Johns Hopkins University are now eyeing future exploits
against other RFID products in the interests of better security, one of the
researchers said this week.  Meanwhile, TI will keep making the compromised
RFID tag in order to meet the needs of applications more sensitive to speed
and pricing than to privacy, according to a TI official.

The Johns Hopkins University Information Security Institute and RSA first
publicized their findings about the RFID security hole in January.  In a
paper posted at, the researchers claim that by cracking
a proprietary cipher, or encryption algorithm in one of TI's DST (digital
signature transponder) RFID tags, they were able to circumvent the tags'
built-in security enough to buy gasoline and turn on a car's ignition.  The
researchers from Johns Hopkins and RSA reverse-engineered and emulated the
40-bit encryption over two months.

  [The full article is on IP:]
IP Archives at:

Sumitomo cyberattack

<Tom Van Vleck <>>
Thu, 24 Mar 2005 16:22:07 -0500

Police foil 220-million-pound 'keyboard hacker' raid on bank,
TimesOnLine, 17 Mar 2005,,2-1529429,00.html

Clinical Healthcare IT, 'error', and safety (Re: Morrell, RISKS-23.79)

<Richard Cook <>>
Thu, 17 Mar 2005 21:24:18 -0600

The note in RISKS-23.79 regarding COPE [Computerized Physician Order Entry
Systems] prompts me to point to our website,, where some of the
papers cited by the JAMA articles by Koppel et al., Garg et al., and
editorial maybe found. The finding that CPOE is a source of new forms of
failure is not surprising.  We have, indeed, predicted this for at least a
decade, as the editorial by Wears & Berg points out. It is not surprising,
either, that some continue to claim that most medical "mistakes" are "caused
by humans".  Although this notion of error has been thoroughly debunked over
the past twenty-five years, the idea is deeply ingrained.

The scientific understanding of the nature of human performance, technology,
and complex systems and their failures traces back to the aftermath of the
Three Mile Island nuclear event in 1979. Woods, Norman, Rasmussen,
Hollnagel, Senders, Moray, Wreathall, and many others spent fifteen years
understanding the relationships between failure and success in domains
including aviation, nuclear power generation, and, more recently, in
healthcare. [There is a useful bibliography in the short paper "A Brief Look
at the New Look in Complex SYstem Failure, Error, and Safety" that can be
found at .] When the patient safety movement began in the
1990's, our hope was that healthcare could avoid getting caught up in the
sterile business of error attribution and counting and quickly move to the
modern view of failure and success. Several scientists, notably David Woods,
spent a great deal of time and effort with groups like the National Patient
Safety Foundation in an effort to 'jump start' healthcare's work on
safety. We achieved only a partial success --- the healthcare world did
'discover' error and become fascinated by it but, after a decade, most of
the leadership now understands that the pursuit of 'error' is unproductive
and a mistaken goal.

The JAMA papers and editorial are correct in their assessment of the current
state of Clinical Healthcare Information Technology (CHIT). What is missing
from the JAMA paper on CPOE and also from the editorial by Wears and Berg is
a clear understanding of why current CPOE is so badly suited to the task of
improving safety. Neither the paper authors nor the editorial writers are
able to look deeply into the design features of these systems or the work
that they are supposed to support. Such close examination reveal, as RISKS
readers will already have anticipated, that it is the failure to produce
USER centered design that is the root cause of the poor performance of
CHIT. The complex activity network that produces patient care is perhaps the
most difficult place to insert interactive computing aids and the designers
of these systems have done little to understand the patterns of work that
occur there or the kinds of support that would be helpful; the paper by
Patterson et al. in J Am Med Informatics Assn 2002:9;540-53 provides a
detailed study. The result is TECHNOLOGY centered systems that generate
failures because they are so ill-suited to the work at hand. Of course the
designers of these things were certain that they were making user centered
designs but the actual results are thoroughly technology centered. As David
Woods said, "the road to technology centered systems is paved with user
centered intentions."

We know, in principle and through demonstration, what it takes to make good
CHIT. As Nemeth, et al., point out in a recent issue of IEEE Systems Man and
Cybernetics (part A, vol 34, 2004), what is needed is detailed, calibrated
understanding of the actual task requirements of the work domain and the
tradeoffs and strategies that workers use to meet these demands. There are
excellent examples of this sort of approach available but , like all good
design, it takes time, money, and more than a little sophistication to do
it. The rush to eliminate "human error" from medicine has led an eager and
somewhat naive group to insist that new CHIT be put in place to forestall
error by practitioners.  Fueled by folk models of human error, this
insistence has produced a whole lot of CHIT that will be the source of a
steady stream of interesting failures over the next decade.

It is unsettling and disappointing to realize that the efforts to produce
really good CHIT are going to require a great deal more time, effort, and
money than has been budgeted. But RISKS readers will recognize that this too
is a common experience with large systems. Many hospitals are already deeply
involved in buying and installing new CHIT and the strong government
pressure to continue this effort is likely to continue. We can only hope
that a parallel effort to understand the technical work of healthcare will
be undertaken so that, in time, it will be possible to make better, more
useful, more user centered technology.

Richard Cook, MD, Cognitive Technologies Laboratory, University of Chicago

Human error and computerized medical systems (Re: Morrell, RISKS-23.79)

<"Don Norman" <>>
Fri, 18 Mar 2005 09:12:58 -0800

In RISKS-23.79, Bob Morrell once again wants to blame the human for error in
complex medical systems. Geesh, I thought that RISKS readers knew better.

Yes, people do make mistakes, but as I and many others have repeatedly
pointed out, in complex systems, there is seldom a single point of failure,
so to trying to assess "the" cause of an error is counterproductive. Yes, it
feels good to be able to blame some person or thing, but this is what I have
called the "blame and train" philosophy. It fails to fix the complex
underlying causes.

If there really is a single point of failure, especially one that repeats
over time, the proper response is to make the system insensitive to this
problem. If we know that a system component is noisy or error-prone (a
transducer, say, or a noisy transmission line), we take care to design the
system so as to be tolerant of those problems. We use error-correcting
codes, or redundancy or we change the procedures. This is frequent with
mechanical and electronic components, but almost never with people. When
people err in this fashion, we punish them, which does nothing to get at the
real cause. We know people transpose digits, confuse complex directions, and
make other well-known and simple errors. Therefore it is a system error not
to have designed the system to be tolerant of these problems.

Morrell gives the following example: "The most common mistake, at its core,
was raw human misunderstanding: conceptual misunderstanding leading to
misinterpretation of medical data (surgeons who thought the higher the
bacterial MIC number, the better the antibiotic, when the reverse is

Gee, what stupid surgeons — at least that is what we are supposed to
believe.  Even this simple example is open to question. These surgeons sound
incompetent: why couldn't they remember that higher MIC numbers are bad?
Well, how many arbitrary little rules do surgeons have to remember? Note
that the human default is that high numbers are good (and that "up" maps to
"higher," "more," "larger," "louder," etc. - all of which usually are
interpreted as "good."  In general, larger numbers mean better (hence all
the jokes about excellent golf and bowling scores). So assuming that high
MIC is better makes sense.

For me to understand whether this was surgeon stupidity or a system problem,
I would ask how many such rules had to be learned, how consistent where
they, and how frequently did this one come into play. Indeed, what is the
meaning of an MIC number? A quick Internet search reveals these two
definitions of MIC (from very different sources):

Definition 1: "The MIC of a drug is defined in broth as the lowest
concentration that prevents visible turbidity of the broth following the
overnight incubation of 105-6 colony forming units (CFU)/ml (obtained during
the log phase of growth)."

Definition 2: "The lowest concentration of antimicrobial agent that inhibits
the growth of the microorganism is the minimal inhibitory concentration
(MIC). The MIC and the zone diameter of inhibition are inversely correlated
(Fig. 10-5). In other words, the more susceptible the microorganism is to
the antimicrobial agent, the lower the MIC and the larger the zone of
inhibition. Conversely, the more resistant the microorganism, the higher the
MIC and the smaller the zone of inhibition."

(I am tempted to say: case closed.  Quick: is high MIC good or bad?  Rule of
thumb: Any definition that has to contain the phrase "in other words" is a
definition in trouble. In this case, after reading the "in other words"
phrase, I still don't know. I think this means that a High MIC number is
good for the organism, but bad for the physician trying to kill it. I still
have no idea of how this translates into the MIC rating for an antibiotic.)

Folks, there are major system errors here. Don't be so quick to blame the
people, even if surface evidence indicts them. The problems are rich,
complex, and deep.  MIC is perhaps a wonderful term for scientists, but it
is a bad term to be used by practitioners.  I sympathize with the
surgeons. We need system thinking, and a deep understanding of the complex
context in which medicine is practiced before we can asses blame and before
we can start to fix the problem, whether with technology or not.

The RISK here is enormous. Well-meaning people claim that technology will
fix the problem of medical errors. Nonsense. Technology is a tool, and
whether it is effective or even more damaging depends upon how it is
deployed. Thinking there is a single source of error - and therefore a
single problem to be solved — will lead us to even worse problems.

Don Norman, Nielsen Norman Group
Northwestern University

Why IE is insecure: flawed logical thinking...

<Craig DeForest <>>
Thu, 24 Mar 2005 09:29:34 -0700

IE appears to be insecure in part because of flawed logical thinking by its
development team.

There is currently a debate of sorts in the news between Mitchell Baker
("chief lizard wrangler" of the Mozilla Foundation) and Dave Massy (head
developer of Internet Explorer) over which web browser is more secure.  In a
recent ZDNET article (also covered on Slashdot; see links at end), Baker
points out that, since IE is tightly coupled ot the Microsoft Windows
operating system, it is bound to be less secure than Mozilla, which is well
separated from its host OS.

Dave Massy's reply is very interesting (link at bottom):
   >The issue of not being part of the OS is an interesting one though that
   >is frequently the subject of misunderstanding.  IE is part of [Microsoft
   >Windows] so that parts of the SO and other applicaaitons [sic] can rely on
   >the functionality and APIs being present.  IE in turn relies on OS
   >functionality to do it's [sic] job.  To be clear there are no OS APIs that
   >IE uses that are not documented on MSDN as part of the platform SDK and
   >available to other browsers and any other software that runs on Windows.

Dave is making a flawed argument:
    - IE uses a documented interface to the OS
    - The OS interface is available to other software on the OS
    - The complexity of our interface is irrelevant to security

The argument is wrong for two reasons: there is a false hidden premise (that
the OS is bulletproof); and the argument itself is invalid (even if the
hidden premise were true, the conclusion would not follow).

One only need read back-issues of RISKS to find case after case of complex,
unanticipated failure modes in complicated interfaces, each element of which
is thought to be secure.  That lesson is at least 30 years old — I am
thinking of the stories about hidden data channels in Multics.

This is of interest to RISKS readers because it is a stunning example of
poor design by flawed logic: even if the IE coding were flawless at the
subroutine level (we can bet that it isn't), Dave's stated attitude toward
interface security would doom it to be susceptible to attack.


Re: Risks of long and short URLs (RISKS-23.80)

<"D.F. Manno" <>>
Thu, 24 Mar 2005 21:32:32 -0500

In RISKS-23.80, Arthur T. writes of a shortcoming with using as
a substitute for typing long URLs, that being that you do not know where a
tinyurl will take you.

The links created by Makeashorterlink <>
first take you to a page displaying the URL that you are to be redirected
to, giving you the opportunity to bail out if you don't want to go there.

D.F. Manno

Risky US Bank Visa product

<"John Meissen" <>>
Thu, 24 Mar 2005 01:14:00 -0800

US Bank has a Visa product targeted at teens (or rather, their parents),
called VisaBuxx. It looks and acts like a standard Visa-logo debit card, but
is more like a prepaid phone card - you pre-load it with value, and it's not
directly tied to any bank account.

Their web site and marketing literature talk about being able to easily add
value to the card by transferring money online from an existing US Bank
checking account. Unfortunately, the system leaves a lot to be desired.

The website has a link for the VisaBuxx program. When you click
on it you're redirected to another site, called  This site is
apparently run by someone called "WildCard Systems". In order to transfer
money from your US Bank checking account to the card you have to provide
WildCard Systems with your checking account number and routing information
and authorization to pull funds from the account, or give them your own
debit card number. While WildCard Systems may be honorable and trustworthy,
the risks in this are so obvious that it's painful. Meanwhile, the Terms Of
Service published on the site go to great lengths to explicitly disavow any
responsibility for anything bad that might result from the use of the site.

The correct way for the bank to have implemented this would have been to
provide the ability to associate the card with your existing Internet
banking identity, and then let you log in through the bank's website and
tell the them to send money from an account to the card rather than allowing
the card operators to pull money from your account.  Having the ability to
provide account data to the VisaBuxx website is useful for non-US Bank
customers, but a legitimate US Bank customer I shouldn't be forced to do it.

I find it mind-boggling that financial corporations still can't see the
obvious when it comes to protecting customer account data. When dealing with
an official bank product I should NEVER have to tell the application
anything about my accounts.

Important PITAC Cybersecurity report released

<Gene Spafford <>>
Fri, 18 Mar 2005 12:29:22 -0500

[As a member of the PITAC and a co-author of the report, I strongly
encourage people will take time to read this and think about how to
help carry out the recommendations.  --spaf]


Vital to the Nation's security and everyday life, the information technology
(IT) infrastructure of the United States is highly vulnerable to disruptive
domestic and international attacks, the President's Information Technology
Advisory Committee (PITAC) argues in a new report.  While existing
technologies can address some IT security vulnerabilities, fundamentally new
approaches are needed to address the more serious structural weaknesses of
the IT infrastructure.

In Cyber Security: A Crisis of Prioritization, PITAC presents four key
findings and recommendations on how the Federal government can foster new
architectures and technologies to secure the Nation's IT infrastructure.
PITAC urges the Government to significantly increase support for fundamental
research in civilian cyber security in 10 priority areas; intensify Federal
efforts to promote the recruitment and retention of cyber security
researchers and students at research universities; increase support for the
rapid transfer of Federally developed cyber security technologies to the
private sector; and strengthen the coordination of Federal cyber security
R&D activities.

To request a copy of this report, please complete the form at, send an e-mail to, or call the
National Coordination Office for Information Technology Research and
Development at (703) 292-4873.  Cyber Security: A Crisis of
Prioritization can also be downloaded as a PDF file by accessing the
link at


The President's Information Technology Advisory Committee (PITAC) is
appointed by the President to provide independent expert advice on
maintaining America's preeminence in advanced information technology.
PITAC members are IT leaders in industry and academia representing the
research, education, and library communities, network providers, and
critical industries, with expertise relevant to critical elements of the
national IT infrastructure such as high-performance computing,
large-scale networking, and high-assurance software and systems design.
The Committee's studies help guide the Administration's efforts to
accelerate the development and adoption of information technologies
vital for American prosperity in the 21st century.

Contact:  "Alan S. Inouye 1-703-292-4540" <>

EEPI - Electronic Entertainment Policy Initiative

<Lauren Weinstein <>>
Fri, 25 Mar 2005 12:38:04 -0800 (PST)

I'm pleased to announce "EEPI" ( ), a new initiative
aimed at fostering cooperation in the areas of electronic entertainment and
its many related issues, problems, and impacts.

I've teamed with 30+ year recording industry veteran Thane Tierney in this
effort to find cooperative solutions to technical, legal, policy, and other
issues relating to the vast and growing range of electronic technologies
that are crucial to the entertainment industry, but that also impact other
industries, interest groups, individuals, and society in major ways.

There are many interested parties, including record labels, film studios,
the RIAA, the MPAA, artists, consumers, intellectual freedom advocates,
broadcasters, manufacturers, legislators, regulators, and a multitude of

The issues cover an enormous gamut from DVDs, CDs, and piracy issues to
multimedia cell phones, from digital video recorders to Internet file
sharing/P2P, from digital TV and the "broadcast flag" to the Digital
Millennium Copyright Act (DMCA) and "fair use" controversies.

Working together, rather than fighting each other, perhaps we can all find
some broadly acceptable paths that will be of benefit to everyone.

For more information, please see the EEPI Web site at:

A moderated public discussion list and an EEPI announcement list
are now available at the site.

Public participation is cordially invited.  Thank you very much.

Lauren Weinstein or or
+1 (818) 225-2800

Please report problems with the web pages to the maintainer