The RISKS Digest
Volume 16 Issue 48

Friday, 21st October 1994

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…

Contents

"The Mother of All Utility Bills."
F. Barry Mulligan
Observed Electro-Magnetic Interference
Henry Troup
Re: Computer risk that nearly proved deadly
Mark Thorson
Gary Koerzendorfer
Cellular Phone Fraud Operator Arrested
Paul Robinson
Chip Maguire
Not enough bytes bites again
Marc Auslander
Inadvertent postal forwarding
V. Michael Bove Jr.
Computer model of Haiti
Phil Agre
Re: Risks of not thinking about what you're stealing
Joel Finkle
Re: Risk of similar interfaces
Chris Norloff
Erann Gat
John Mainwaring
Re: Software reuse
David Honig
Re: Greyhound
Danny Burstein
CNID and Don Norman — CNID can be private
Justin Wells
Re: CNID
Andrew Klossner
Phil Agre
Info on RISKS (comp.risks)

"The Mother of All Utility Bills."

"F. Barry Mulligan" <MULLIGAN@ACM.ORG>
Fri, 21 Oct 1994 13:08:57 -0500 (CDT)
From The Atlanta Constitution, Tues 18 Oct 1994, p.1, by Christopher C. Warren

  Imagine a single monthly statement listing all utility charges, including
  phone, cable, gas, electricity, water, garbage collection and sewerage
  charges.  It could be the mother of all utility bills and would allow
  consumers to write only a single check for all their services.  One Check,
  as the proposal is being touted, would ease consumer's household management
  by reducing utility bills to one monthly payment, said Maureen Bailey, vice
  president of public affairs with American Express, the company proposing the
  service.

    The article goes on to describe the pilot test being proposed for the
Atlanta metro area. The cost of the service would be shared by the utilities
and the consumer.

    Risks?  A little late with one payment and you're instantly in arrears
with every company in town. Billing disputes "still would be handled through
the individual utility companies", but what if the utility says it didn't
get a payment you sent to the service company?  If your combined statement is
mailed on the 15th and a utility transmits a new charge to the service bureau
on the 16th, what happens to the payment grace period?  If you've ever had to
rob Peter to pay Paul, how do you deal with Peter & Paul, Amalgamated?
    Perhaps the real question is 'Do I want to give a complete, itemized
description of all monthly utility consumption to American Express?' (and pay
for the privilege).


Observed Electro-Magnetic Interference

"henry (h.w.) troup" <hwt@bnr.ca>
Fri, 21 Oct 1994 13:39:00 -0400
On Thursday 20 October 1994, I attended an Emergency Response seminar held in
the auditorium of a high-tech institute. There were a number of fire and
ambulance personnel there "in-service" with two-way radios. A radio
transmission inside the auditorium triggered all kinds of equipment - the
video projector turned on, the slide projector turned on, and some
unidentified motor noise started up.

Nothing new, but an example of a widespread problem (still sometimes not
accepted as real.)

Henry Troup - H.Troup@BNR.CA (Canada)


Re: Computer risk that nearly proved deadly (Maniscalco, RISKS-16.47)

Mark Thorson <eee@netcom.com>
Fri, 21 Oct 1994 10:31:02 -0700
The posting in RISKS-16.47 about reprogramming pacemakers through audio
tones seems a bit alarming to me.  Does that mean someone could have a
tape player (i.e. "boom box") with a tape of pacemaker programming signals
which would be deadly to certain members of the public?  Imagine some guy
walking around with that on a crowded train or bus.

How about cutting into the SCA-encoded Muzak played in supermarkets?
That's broadcast on the same frequencies used for ordinary FM radio.
Often, wireless mikes are used at lectures and concerts.  Care to find
out how many Rolling Stones fans have pacemakers?  The power level
from the speakers at a concert hall ought to be much more than is needed
for programming.

Anyone know the name of the company that made the malfunctioning pacemaker?
It might be interesting to do a patent search on them.

Mark Thorson  (was mmm@cup.portal.com, but now eee@netcom.com)


Re: cauterizer corrupts pacemaker

Gary Koerzendorfer <garyk@hpesds3.cup.hp.com>
Fri, 21 Oct 94 16:20:49 PDT
I was surprised that a pacemaker would be susceptible to this interference,
that it didn't figure it out, and apparently had no fallback mode. I'm
not sure what EMI field strength the electrocauterizer created; should
a patient be advised to avoid proximity to arc-welders and to 50KW
broadcasting facilities like the one next to a bridge in San Francisco Bay?

Secondly, a state-verification circuit (or even a "heartbeat"!) could
have detected the malfunction, and regressed to a fallback fixed rate.
Your own heart has a 30 beats/minute tertiary protective function -
you're unable to engage in strenuous activity, but you remain alive
until you can seek help.

An incapacitated patient would be unable to inform ER staff that they
had a pacemaker - perhaps if I had one I'd wear a MedicAlert tag.

Gary Koerzendorfer, Hewlett-Packard Co., Systems Techn. Div., 19447 Pruneridge
Ave. m/s 42L2, Cupertino Calif  95014  garyk@cup.hp.com  (408) 447-4783

   [For younger RISKS readers, TWO cases of deaths due to pacemakers being
   affected by interference are included in the RISKS archives.  PGN]


Cellular Phone Fraud Operator Arrested (Summary and Comments)

Paul Robinson <PAUL@tdr.com>
Thu, 20 Oct 1994 21:37:55 -0500 (EST)
The following article summary is followed by some comments:

   In a Front Page article appearing in the Wed 19 Oct 1994 {Washington (DC)
Times} entitled "High-Tech sleuthing busts cellular phone fraud ring" reporter
Doug Abrahms tells us that Clinton Watson and two other persons were arrested
Monday for selling cellular phones with altered serial numbers, causing the
charges to be sent to legitimate cellular users.
   According to an Indictment in U.S. District Court in San Jose, when police
raided Watson's house, they found 30 phones with counterfeit ID, 16 altered
memory chips and 600 mobile phone numbers which could be used for fraudulent
calls.  Some of Mr. Watson's phones had as many as 12 different ID numbers,
thus spreading usage patterns over a large area.  Other phones were designed
to allow the ID to be changed at will.
   Police and cellular companies have turned to using
more sophisticated means to find illegal cellular phones, including
helicopters, voice prints and traffic analysis.
   Mr. Watson is a Computer Programmer who designed his own software to
program integrated circuits to include numbers read from scanners used on the
cellular band.  The phones so set up were referred to as "lifetime" phones
since they never got a bill.  They sold for $1,200 to $1,500 and have been
found all over North America, according to Ron Nessen of the Cellular
Telecommunications Industry Association (CTIA), which estimates that cellular
fraud is a $1 million a day problem, with people stealing cellular IDs by
waiting near tunnels, airports and parking lots to snatch the ID code
transmitted.
   New York's NYNEX is introducing a PIN code on cellular calls.  The Mayor
and Police Commissioner of New York City have had the IDs for their cellular
phones stolen six times this year.  A division of TRW is developing a means to
prevent calls unless the user's voice print matches the print on file.

Comments:

1.  Cellular Companies have been notorious for evading security problems in
their phones.  Rather than spend the money to add encryption in their switch
software, they got a law passed to make it illegal to listen to cellular
frequencies and to build equipment that can monitor cellular bands.

2.  Cellular phones transmit call information in the clear, so a thief can
just use someone else's number and steal a few minutes of airtime from them;
if you bleed 10,000 customers of ten extra minutes a month, almost none of
them individually will recognize that their bill is ten minutes too high.
Unless customers complain, the Cellular Company won't care.

3.  A typical practice of an aerospace/military contracting company like TRW
is to try an implement and expensive complicated system such as voice print
matching instead of something simple and cheap like a device to implement
either Kerberos validation, S/Key style one-time passwords, or MD-4/MD-5
arithmetic checksum of some stored value.  Putting such methods in as an
inexpensive box like a Radio Shack tone dialer might cost users $20 and
installing it in new phones might cost an extra $2 or $3.  Persons having
portable PCs could run a program to generate the code.  Since everything is
done without a secret being transferred, the software to do this can be public
and nothing is compromised.

4.  Does using a biometric validation system on a communications network
scare anyone?  I can think of a half-dozen reasons to dislike it, including:
- use of the system to track and locate dissidents and anyone the people
  who run the government don't like;
- my sister wants me to call someone for her and find out something
  without them knowing it's her asking; I don't match her car phone profile;
- I borrow her car to do an errand; I can't call her back to let her know
  what I found out for her;
- Bugs in the software might not recognize the owner with a cold, after an
  accident that damages their throat, or after some forms of surgery;
- Checking voice prints will require very heavy processing capability,
  quite likely slowing down call connection times;
- I bug someone's car and simply play back the recording to unlock their
  phone.

I think that this is an attempt to "kill flies with nuclear weapons," e.g.
excessive overkill.  There are cheaper alternatives such as mathematical
verification that will probably be quite effective without using a system that
requires expensive and complicated subsystems such as voice print recognition.


Re: (mobile-ip) Cellular Phone Fraud Operator Arrested

<maguire@it.kth.se>
Sat, 22 Oct 94 01:00:54 +0100
Actually - FBI/Cellular operators/... could just buy suitable equipment from a
major electronics manufacturer in France - it reportedly even includes
key-word spotting.

Certainly why I look forward to widely available portable crypto for mobile
voice and data.

Chip


Not enough bytes bites again

Marc Auslander <marc@watson.ibm.com>
Fri, 21 Oct 1994 13:41:16 -0400
---222---
NOTICE ADVISORY TO NAVSTAR USERS (NANU) 222-94266
SUBJ: TIME TRANSFER DATA ANOMALY
1. CONDITION: USERS OF GPS TIME TRANSFER INFORMATION ARE ADVISED
THAT THE DATA IS SUSPECT AS OF DAY 266 (23 SEP 94) BEGINNING AT
1500 UTC AND CONTINUING UNTIL FURTHER NOTICE.
2. GPS NAVIGATION POSITIONAL DATA ACCURACY IS UNAFFECTED.
3. POC: CPT JAMES AT COMMERCIAL (719) 550-6378 OR DSN 560-6378.

---223---
NOTICE ADVISORY TO NAVSTAR USERS (NANU) 223-94266
SUBJ: USERSET TIME TRANSFER ERRORS
1. CONDITION: INVESTIGATION HAS REVEALED THERE IS NO CURRENT OR
PAST PROBLEM CONCERNING TIME TRANSFER INFORMATION. WE HAVE DETERMINED
THAT SOME USERSETS MAY NOT ACCOUNT FOR ROLLOVER OF THE 8 BIT GPS UTC
REFERENCE WEEK. IAW ICD GPS 200, PARAGRAPH 20.3.3.5.2.4, PAGE 18 OF
SUBFRAME 4 INCLUDES THE PARAMETERS NEEDED TO RELATE GPS TIME TO UTC.
THE USER MUST ACCOUNT FOR THE TRUNCATED NATURE OF THE UTC REFERENCE
WEEK. THIS USERSET DISCREPANCY WILL OCCUR EVERY 255 WEEKS AND WILL
LIKELY CLEAR AT END OF WEEK ROLLOVER.
2. POC: 1LT PETERSON AT COMMERCIAL (719) 550-6378 OR DSN 560-6378.

Marc Auslander   <marc@watson.ibm.com>   914 784-6699  (Tieline 863 Fax x6306)


inadvertent postal forwarding

V. Michael Bove Jr. <vmb@media.mit.edu>
Fri, 21 Oct 94 16:12:43 -0400
The following concerns an acquaintance whom I'll call S, who works at the
corporate headquarters of a large firm which I'll call FooCorp (the pseudonyms
are used because none of what happened seems to be FooCorp's fault, but some
of their clients might be a bit distressed to hear about it).

Having recently changed residences, S filed a mail-forwarding order in her
name at her former post office.  A few days later, mail addressed to the
offices of a number of FooCorp employees started showing up at her new house,
complete with yellow computer-generated forwarding stickers.  How this
happened is partial conjecture on S's and my part, as the postmaster in S's
former town of residence wasn't terribly helpful in unraveling the mystery.
The trigger seems to be the fact that as part of her employment S has a
commercial credit card in the name of FooCorp, and the bills were mailed to
``FooCorp, [S's old home address].''  Somehow, some combination of postal
employees and postal software decided therefore to generate a forwarding order
such that mail addressed to FooCorp passing through the postal sorting
facility nearest to S's former home should be forwarded to S's new address.

The outcome is that various important-looking pieces of FooCorp's mail were
diverted to a rural mailbox in front of a horse barn in the middle of nowhere,
and apparently no one in the Postal Service thought this at all odd.

The postmaster thinks the forwarding order has now been eradicated, and S has
contacted her credit card issuer to cause the bills to be mailed in her name,
not FooCorp's.  However, the misdelivered mail hasn't totally stopped, since
while the order was in effect, the Postal Service dutifully notified anyone
whose envelope said ``Address Correction Requested'' that FooCorp had moved;
there are now at least a few mailing databases out there which think that
FooCorp is located at S's house.

The lesson, of course, is that procedures that try automatically to do the
right thing without being asked sometimes need a reality check applied to the
results.


Computer model of Haiti

Phil Agre <pagre@weber.ucsd.edu>
Thu, 20 Oct 1994 18:15:37 -0700
In an article in The Nation, Allan Nairn asserts that the American military
force occupying Haiti is more concerned about the popular movement that
elected Jean-Bertrand Aristide than it is about the oligarchy that created
the attaches.  I don't know whether this assertion is fair, but Risks readers
might be interested in one part of his evidence.  The full citation is:

  Allan Nairn, The eagle is landing, The Nation 259(10), 3 October 1994,
  pages 344-348.

Here's a quote:

  ... the Pentagon's Atlantic Command (ACOM) has commissioned Booz, Allen,
  Hamilton, a corporate consulting firm, to devise a computer model of Haitian
  society.  A similar model was ordered for Iraq for Desert Storm.  The model
  tries to predict "the effects of social, political and economic actions on
  various sectors of society".  In an April 29 report Booz, Allen presented a
  "Power Relationship Matrix" which divides Haitian society into seven groups,
  including the "Lower Class Majority", and asks questions like "What would
  mobilize the masses to take action?"

  The crux of the Booz, Allen/ACOM planning theory is thus: "Whether political
  power is a direct function of popular support or based on the allegiance
  of key groups and coercion of the remainder of the populace, cohesion of
  support is a critical question in assessing political power".  They place
  greatest emphasis on the importance of "Organized Civil Society" — popular
  and professional groups, unions and associations, development workers --
  seeking to identify the points at which mass cohesion will crack.  This,
  they say, is the key to any program for "control of the populace".

  Their priority is to build an "organized information bank" and to run a
  systematic, ongoing "assessment of the relative strengths of opposition
  organizations", as well as of leading "political personalities".  "The
  tracking of opposition organizations", they say, "should be limited to those
  which are known to have a basis of political action and some established
  capacity for taking political action".

  As the Washington Office on Haiti has documented in detailed reports, A.I.D.
  [the US Agency for International Development] is already exploring this
  divide-and-conquer strategy in Haiti, seeking to cultivate and fund, as on
  embassy memo put it, "responsible elements within the popular movement"
  along with "moderate Duvalierist factions".  (pages 347-348)

An exercise for the reader: if they implemented such a program for your home
town, what would it look like?

Phil Agre, UCSD


re: Risks of not thinking about what you're stealing

Joel Finkle <jjfink@skcla.monsanto.com>
Fri, 21 Oct 1994 12:55:34 -0500
Tumbling a beeper's activation code is trivial.  You can take your beeper to
any beeper service company, ask to be switched to their line of service, and
they'll recode and activate you, assuming (a) that you're using the same
brands that they are and (b) you're *not* using the same service company they
are (or they can't switch you).  Presumably, a stolen beeper could be handled
as if you owned it, then there's a high likelyhood that you'd get it
reactivated.


Re: Risk of seeming similar interfaces (Elkins, RISKS-16.47)

<cnorloff@tecnet1.jcte.jcs.mil>
Thu, 20 Oct 94 22:48:41 EDT
ARRRRGGGHH!!  And the risk of idiots designing switches!  It does not take a
rocket scientist to figure out that a critical switch should be put in a
hard-to-access location or should have a protective cover over it.  This kind
of stuff is taught in very basic human engineering classes.

There are certain errors designers are doomed to keep repeating, and switch
design is one of them.

Chris Norloff  cnorloff@tecnet1.jcte.jcs.mil


Re: Risk of seeming similar interfaces (Elkins, RISKS-16.47)

Erann Gat <gat@aig.jpl.nasa.gov>
Fri, 21 Oct 94 11:09:00 PDT
Monta Elkins writes:

>on the Power Mac this is the POWER button

Geez!  Will Apple never learn?  The original Apple II was notorious for having
its reset button on the keyboard immediately adjacent to the return key, with
predictable results.  If any company should know to put power and reset
buttons far away from everything else on the machine you'd think it would be
Apple.  Sigh.

Erann Gat  gat@robotics.jpl.nasa.gov


Re: Risk of seeming similar interfaces (Elkins, RISKS-16.47)

"john (j.g.) mainwaring" <crm312a@bnr.ca>
Fri, 21 Oct 1994 14:33:00 -0400
The late lamented Commodore company did the same with the Amiga 3000, which
has a push power switch in almost the same location as the disk eject button
of the A1000.  I don't suppose that was the only reason for the company's
demise.  Still, fond as I have been of the A3000 since the day I got it, that
design feature led me to wish an even worse fate on the company on numerous
occasions during my first several months of ownership.


Re: Software reuse (Gonzales, RISKS-16.47)

David Honig <honig@buckaroo.ICS.UCI.EDU>
Fri, 21 Oct 1994 12:18:24 -0700
>Published reusable software should be as carefully designed and verified as
>any other safety critical software, since its publishers will have no control
>over what systems it gets reused in.

I think this is wrong.  Ensuring safety is the requirement of the human
organization producing a potentially hazardous product.  The *organization*
must have adequate quality checks.  One engineer's bad day, one alpha
particle, or one bolt's failure, should not harm people.

One wouldn't disallow the reuse of used power supply designs, for instance.
Instead, if you were putting them into an important machine, you would not
only test their published performance, but perhaps *add* additional
safety-relevant tests.  And then you would design in several of them if the
application justified it.

If you want to reuse some code (or other artifact) in a RISKy application,
maybe you'll have to characterize it more than a less critical user.  What
else is new?

Computer professionals should learn that (organizations of) people are
responsible; technical professionals are responsible for making technical
observations and their implications known to others in the organization.


Re: Greyhound (RISKS-16.47)

danny burstein <dannyb@panix.com>
Thu, 20 Oct 1994 22:43:44 -0400 (EDT)
While I haven't seen the original piece, I would expect that it glossed over
what may very well be a major, if not -the- major, cause of of the problems --
namely, the horrendous management-labor problems at the company that led to a
strike/lockout/mass firing (take your choice....) of the vast majority of bus
operators.  The bosses, though, as well as the finance MBA types, stayed in
office. And, as the article mentioned, cheerfully exercised their stock
options to make huge profits.

  -dannyb@panix.com (or dburstein@mcimail.com)


CNID and Don Norman — CNID can be private

Justin Wells <rjwells@undergrad.math.uwaterloo.ca>
Fri, 21 Oct 1994 13:29:49 -0400
In Don Norman, in his book "Things That Make Us Smart", makes some interesting
points about Call Number ID.  He writes that the emphasis on NUMBERS stems
from a mechanistic view of the phone system — a human view of the phone
system wouldn't depend so heavily on numbers (which we have trouble
remembering).

He suggests that other information should be sent instead of a telephone
number.  Perhaps each telephone owner could set a string that accompanies
their call instead of their phone number.  It need not be more than a few
characters — maybe your first name, maybe the name of your company.

His point is that the average person wants to know WHO is calling, and only
businesses, etc., really care about the phone number.  This solution should
retain the value of CNID (knowing whether you want to answer the phone), but
eliminate the privacy risks (including the need for call blocking, etc.)

I found lots of excellent commentary on this and other sorts of RISKS in his
book.  I know he's posted here before, and I noticed he has a reference to
RISKS in his book.  I recommend it to everyone.

Justin

   [Don also wrote a column for the April 1993 CACM Inside Risks, which
   has been adapted as a section in my COMPUTER-RELATED RISKS book.  PGN]


Re: CNID (Preece, RISKS-16.47)

Andrew Klossner <andrew@frip.wv.tek.com>
Fri, 21 Oct 94 11:54:46 PDT
  Require the phone company to assign a second unique number ... to each
  telephone... I don't think this represents a serious  technological hurdle...

It does.  Almost all area codes in North America are more than half populated.
In order to double the number of phone numbers, most area codes would have to
be split.

  -=- Andrew Klossner  (andrew@frip.wv.tek.com)


Re: CNID

Phil Agre <pagre@weber.ucsd.edu>
Thu, 20 Oct 1994 19:09:09 -0700
In his comment (RISKS-16.47) on my message about CNID, Peter da Silva
<peter@nmti.com> refers to me as an "Anti-CNID fella".  But I am not opposed
to CNID, only to bad implementations of it that make free choices about CNID
difficult.  At the same time, he is correct to criticize my generalizations
about "CNID proponents".  I didn't mean to refer to everyone who finds CNID
useful, only the folks who wish to profit from bad implementations of CNID
and can hire lobbyists for this purpose.

Phil Agre

Please report problems with the web pages to the maintainer

x
Top