The Risks Digest

The RISKS Digest

Forum on Risks to the Public in Computers and Related Systems

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

Volume 30 Issue 96

Wednesday 12 December 2018


A note on submissions to RISKS
The War on Truth Spreads
Annoyed Baltimore Drivers Want City To Crack Down On 'Squeegee Kids'
Your apps know where you were last night, and they're not keeping it secret
The 'Weird Events' That Make Machines Hallucinate
Linda Geddes
Barclays customers can now 'switch off' spending
Ships infected with ransomware, USB malware, worms
Catalin Cimpanu
Taylor Swift tracked stalkers with facial recognition tech at her concert
The Verge
What Happens When You Reply All to 22,000 State Workers[?]
U.S. border officers don't always delete collected traveler data
Marriott Data Breach Is Traced to Chinese Hackers as U.S. Readies Crackdown on Beijing
Starwood Hotels
PGN via Mabry Tyson
Why I'm done with Chrome / A Few Thoughts on Cryptographic Engineering
Cryptography Engineering
Screen Time Changes Structure of Kids' Brains: Groundbreaking study
Re: Teen electrocuted while using headphones on plugged-in mobile phone
Richard M Stein
Re: Toronto auto theft ...
Steve Lamont
Re: Rudy Giuliani Says Twitter Sabotaged His Tweet
Amos Shapir
Info on RISKS (comp.risks)

A note on submissions to RISKS

"Peter G. Neumann" <>
Mon, 10 Dec 2018 11:11:14 PST
                         - BEGIN RANT -

OK, RISKS readers, “I'm mad as hell, and I'm not going to take it any
more.''  I'm really fed up with trying to edit what some of you send me,
trying to produce nice clean readable issues of RISKS, without errors.  I'm
not giving up on putting out RISKS issues, but the time it takes to put out
each issue has recently been escalating.  Please don't bother to complain
about characters that are garbled.  It's wasting your time.  I'm not

From the very early RISKS issues in 1985, I have expressed a desire to
receive messages with ASCII characters; later on, I made a plea to
completely avoid attachments in Word, pdf, html, or even encoded ASCII.  I
process RISKS e-mail with an archaic ASCII-happy mail system, because it
hugely simplifies my ability to delete more than 80% of the incoming mail
sight unseen (lots of spam), and then trying to cull out and lightly edit
your *good* contributions.  Nevertheless, I still get smart quotes and smart
apostrophes from Mac users, encodings of spaces as underscores (or some
weird unprintable character) and equal signs from Windows systems that
insist on encoding certain ASCII characters as non-ascii characters, rampant
=E2=80 encodings, long lines split with an equal sign at the end of each
line, non-ASCII From: addresses (e.g., from Mateo), copies of entire RISKS
issues as attachments when you are responding to an item in a previous
issue, the entire ASCII text of your would-be contributions completely
duplicated in horribly fulsome html, rampant extra junk appended (from
Richard Stein *), URLs that come out with %3A%2F%2F encodings, and more.
UTF-8 might help a little, but is primarily useful for attachments that use
it consistently.  Then, for your ease of reading, I try to unscramble overly
long URLs and verify my attempts at creating shorter ones, and remove all
the extra cruft created by Office-365-safelinks URL enscramblings that
evidently offer no real security anyway.  Furthermore, I do not have time to
cope with alternative approaches, such as your putting jpeg files on your
website for me to view with a browser.

Perhaps needless to say, I would greatly appreciate if you can spend just a
few more moments in your submissions to have a little more concern for my
own well-being.  ASCII is ASCII, and emacs is emacs, and I will remain a
troglodyte in order to continue to moderate RISKS for you.  I am sorry that
I do not readily handle all of your special characters.  Clearly, if RISKS
had to deal with submissions in Cyrillic, Kanji, Farsi, Arabic or whatever,
I would have to do things very differently—or simply completely give up
running a seriously moderated digested new group (where you can create your
own undigestifier if you prefer).  However, if you think you have a better
solution, please let me know.  THANKS in advance for your consideration.

                         - END RANT -

[* Footnote from each of Richard Stein's contributions in this issue:
  ad finitum—for 77 lines of similar meaningless garbage.

  Let's see who gags on this issue, where I have intentionally left in
  a few outliers.

The War on Truth Spreads (NYTimes)

"Peter G. Neumann" <>
Mon, 10 Dec 2018 12:33:42 PST
An editorial with the above caption in the 10 Dec 2018 issue of *The New
York Times* considers systemic incursions on freedom of the news media
around the world, including the Philippines. Hungary. Saudi Arabia. Turkey,
China, Russia. and even the U.S.  Internet censorship and Internet misuse
have both played significant roles.  In short, we have vastly transcended
even the horrors of George Orwell's *1984*.

Annoyed Baltimore Drivers Want City To Crack Down On 'Squeegee Kids' (

Richard Stein <>
Mon, 10 Dec 2018 10:39:01 +0800

How will an autonomous vehicle will address a squeegee bum assault? A horn
toot? Redirection of windshield sprayers?

Your apps know where you were last night, and they're not keeping it secret (NYTimes)

geoff goodfellow <>
Mon, 10 Dec 2018 08:55:07 -1000
Every moment of every day, mobile phone apps collect detailed location
data.Data reviewed by The New York Times shows over 235 million locations
captured from more than 1.2 million unique devices during a three-day period
in 2017.

Dozens of companies use smartphone locations to help advertisers and even
hedge funds. They say it's anonymous, but the data shows how personal it is.


The millions of dots on the map trace highways, side streets and bike trails
-- each one following the path of an anonymous cellphone user.

One path tracks someone from a home outside Newark to a nearby Planned
Parenthood, remaining there for more than an hour. Another represents a
person who travels with the mayor of New York during the day and returns to
Long Island at night.

Yet another leaves a house in upstate New York at 7 a.m. and travels to a
middle school 14 miles away, staying until late afternoon each school day.
Only one person makes that trip: Lisa Magrin, a 46-year-old math teacher.
Her smartphone goes with her.

An app on the device gathered her location information, which was then sold
without her knowledge. It recorded her whereabouts as often as every two
seconds, according to a database of more than a million phones in the New
York area that was reviewed by The New York Times. While Ms. Magrin's
identity was not disclosed in those records, The Times was able to easily
connect her to that dot...


The 'Weird Events' That Make Machines Hallucinate (Linda Geddes)

ACM TechNews <>
Mon, 10 Dec 2018 11:36:58 -0500
Linda Geddes, BBC News, 5 Dec 2018 via ACM TechNews, 10 Dec 2018

Computers can be tricked into misidentifying objects and sounds, raising
issues about the real-world use of artificial intelligence (AI); experts
call such glitches `adversarial examples' or `weird events'.  Said the
Massachusetts Institute of Technology (MIT)'s Anish Athalye, “We can think
of them as inputs that we expect the network to process in one way, but the
machine does something unexpected upon seeing that input.''  In one
experiment, Athalye's team slightly modified the texture and coloring of
certain physical objects to fool machine learning AI into thinking they were
something else. MIT's Aleksander Madry said the problem may be rooted partly
in the tendency to engineer machine learning frameworks to optimize their
performance on average. Neural networks might be fortified against outliers
by feeding them more challenging examples of whatever scientists are trying
to teach them.

Barclays customers can now 'switch off' spending (

Richard Stein <>
Tue, 11 Dec 2018 13:13:05 +0800

“The idea is to help vulnerable customers, particularly problem gamblers, or
those in serious debt.''

Cellphones, while generally indispensable for communication purposes, are
gateway devices that can enable addictive behaviors. A compulsive gambler
smart enough to configure a cellphone application should recognize that
professional counseling and therapy is more effective than a voluntary, and
easily overridden, videogame context configuration setting.

A flick of the cellphone application switch precludes a bank debt card from
being used for problematic and harmful purposes at certain `classes' of
vendors: “Groceries and supermarkets, restaurants, takeaways, pubs and bars,
petrol stations, gambling - including websites, betting shops and lottery
tickets, premium rate websites and phone lines, including TV voting,
competitions and adult services.''

Risk: Financial/lifestyle surveillance and profile disclosure via data
breach or explicit sale.

That a financial institution, not widely known for their altruism, promotes
this application implies that an intimate profile of an addict as customer
arises from consolidated spending patterns. Difficult to assess how this
business intelligence might be exploited internally, or by a 3rd party if
terms of service stipulate sale and reuse conditions.

Ships infected with ransomware, USB malware, worms (Catalin Cimpanu)

Gene Wirchenko <>
Wed, 12 Dec 2018 11:38:44 -0800
Catalin Cimpanu for Zero Day, 12 Dec 2018

Ships infected with ransomware, USB malware, worms
Ships are the victims of cyber-security incidents more often than people
think. Industry groups publish cyber-security guidelines to address issues.

selected text:

For example, the guidelines include the case of a mysterious virus infection
of the Electronic Chart Display and Information System (ECDIS) that ships
use for sailing.

     A new-build dry bulk ship was delayed from sailing for several days
because its ECDIS was infected by a virus. The ship was designed for
paperless navigation and was not carrying paper charts.

       [No backup!]

Ships were also impacted by ransomware, sometimes directly, while in other
incidents the ransomware hit backend systems and servers used by ships
already in their voyage at sea.

For example, in an incident detailed in the report, a shipowner reported not
one, but two ransomware infections, both occurring due to partners, and not
necessarily because of the ship's crew.

      [And there are other examples given.]

Taylor Swift tracked stalkers with facial recognition tech at her concert (The Verge)

=?utf-8?Q?Jos=C3=A9=20Mar=C3=ADa=20Mateos?= <>
Wed, 12 Dec 2018 15:13:09 -0500

Taylor Swift held a concert at California's Rose Bowl this past May that was
monitored by a facial recognition system. The system's target? Hundreds of
Swift's stalkers.

Swift's facial recognition system was built into a kiosk that displayed
highlights of her rehearsals, which would secretly record onlookers' faces.
According to Rolling Stone, which spoke with a concert security expert who
observed the kiosk, attendees who looked at the kiosk were immediately scanned.
Afterward, the data was sent to a `command post' in Nashville, Tennessee that
attempted to match hundreds of images to a database of her known stalkers.

José María (Chema) Mateos

What Happens When You Reply All to 22,000 State Workers[?] (NYTimes)

Monty Solomon <>
Tue, 11 Dec 2018 01:26:32 -0500

Reply All, the scourge that has afflicted office workers everywhere, has hit
22,000 government employees in Utah.

U.S. border officers don't always delete collected traveler data (

Richard Stein <>
Wed, 12 Dec 2018 16:39:58 +0800

“Privacy advocates aren't just concerned about warrantless device searches
at the border because of the potential for deliberate abuse—it's that the
officials might be reckless. And unfortunately, there's evidence this is the
case in the U.S. Homeland Security's Office of the Inspector General has
released audit findings showing that Customs and Border Protection (CBP)
officers didn't properly follow data handling procedures in numerous
instances, increasing the chances for data leaks and hurting

Assembled and maintained by CBP, this honeypot of mobile device contacts,
photos, downloads, browser history, call logs, and credit card/app profiles
will likely attract ex-filtration attempts.

A comprehensive repository of personal data that can be correlated against
many other dark-net sources, and maliciously exploited for profit or
criminal intent.

Marriott Data Breach Is Traced to Chinese Hackers as U.S. Readies Crackdown on Beijing (NYTimes)

Monty Solomon <>
Wed, 12 Dec 2018 10:07:20 -0500
Marriott Data Breach Is Traced to Chinese Hackers as U.S. Readies Crackdown
on Beijing

The Trump administration is expected to indict hackers and roll out import
restrictions out of concern that Beijing will not easily change its trade,
cyber[security? privacy? ...] and economic practices.

Starwood Hotels

"Peter G. Neumann" <>
Wed, 12 Dec 2018 16:19:45 -0800
  [Thanks to Mabry Tyson.]

21 Nov 2015 (a year or so after the initiation of the intrusion currently in
the news)

  Starwood Hotels and Resorts Worldwide Inc. is the latest known hotel
  target of cyber-attackers. The company on Friday announced that hackers
  had injected malware into point of sale systems at some of its hotels in
  North America.

  That malware ultimately made it possible for unauthorized parties to tap
  into the payment card data of some hotel guests. Starwood, which operates
  brands including Four Points by Sheraton, Aloft, Element, and Westin, now
  joins the *Trump Hotel Collection and the Hilton chain* of hotels on the
  list of hotel data breaches.

  As soon as it discovered the breach, Starwood hired outside forensics
  experts to investigate the depth and breadth of the attack. The result:
  investigators discovered malware installed in the point of sale systems of
  some of its restaurants, gift shops and other systems.  *The company said,
  at this time it doesn't appear Starwood's guest reservation or preferred
  guest membership systems were breached.*

  “Starwood certainly isn't the first company to be affected by point of
  sale malware. The path from discovery to recovery is well-worn at this
  point.  In some cases this malware has been present for *more than a
  year.*'' While the incident may seem like a point in time, it's really a
  lengthy campaign of data theft, Erlin said, adding that he's surprised
  that fraudulent activity from stolen card data wasn't discovered sooner.

Incidentally, a better reference on the 2015 MARRIOTT intrusion (which
started July 2014, and ended April 2015) is this (which refers to an earlier
malware incident in 2014):

Why I'm done with Chrome / A Few Thoughts on Cryptographic Engineering (Cryptography Engineering)

Dan Jacobson <>
Wed, 12 Dec 2018 02:45:00 +0800

“One argument is that Google already spies on you via cookies and its
pervasive advertising network and partnerships, so what's the big deal if
they force your browser into a logged-in state? One individual I respect
described the Chrome change as `making you wear two name tags instead of

Screen Time Changes Structure of Kids' Brains: Groundbreaking study (Bloomberg)

the keyboard of geoff goodfellow <>
Sun, 9 Dec 2018 16:13:57 -1000
Smartphones, tablets and video games are physically changing the brains of
adolescents, early results from an ongoing $300 million study funded by the
National Institute of Health have shown, according to a report by *60

Scientists will follow more than 11,000 nine- to 10-year-olds for a decade
to see how childhood experiences impact the brain and affect emotional
development and mental health. The first bits of data suggest that the
onslaught of tech screens has been transformative for young people—and
maybe not for the better.

In brain scans of 4,500 children, daily screen usage of more than seven
hours showed premature thinning of the brain cortex, the outermost layer
that processes information from the physical world. Though the difference
was significant from participants who spent less screen time, NIH study
director Gaya Dowling cautioned against drawing a conclusion.  “We don't
know if it's being caused by the screen time. We don't know if it's a bad
thing.  It won't be until we follow them over time that we will see if there
are outcomes that are associated with the differences that we're seeing in
this single snapshot.''  (according to an advance script)

Early results from the study, called Adolescent Brain Cognitive Development
(ABCD), have determined that children who spend more than two hours of
daily screen time score lower on thinking and language tests. A major data
release is scheduled for early 2019...
YOU CAN VIEW the (~13 min) segment here:

Re: Teen electrocuted while using headphones on plugged-in mobile phone (Lesher, RISKS-30.95)

Richard M Stein <>
Sun, 9 Dec 2018 16:37:24 +0800
[It is not] surprising to learn about counterfeit chargers and phony
qualification labels that certify safety. Not many consumers can distinguish
real labels from fake, nor are they inclined when price often determines
purchase motive. Similar problem for pharmaceuticals, auto parts, and
aircraft parts. Makes you wonder about drug and travel safety given forgery
incident frequency. Thx.

Re: Toronto auto theft ... (RISKS-30.95)

Steve Lamont <>
Tue, 11 Dec 2018 14:43:59 -0800
You will note if you read the story that no one has produced an actual relay
device in evidence.  The rather murky surveillance video still shows the
alleged miscreant carrying. . . something but whether it's a fob repeater or
just a plastic bag containing standard burglar tools is entirely unclear to

Until I see an actual device, color me skeptical.

Re: Rudy Giuliani Says Twitter Sabotaged His Tweet (RISKS-30.95)

Amos Shapir <>
Mon, 10 Dec 2018 09:43:10 +0200
Actually this *is* Twitter's fault!  (Though not in the way Giuliani
thinks).  It is obvious that Giuliani was not aware that Twitter is turning
periods in his post into links.  But did Twitter do anything to make their
users—especially the less technically inclined—aware of this fact?  Is
there a way to turn this mis-feature off?  Why did Twitter make it active by
default, and in such a dumb way (the generated link was not valid as
written, so it's obvious the user did not intend to enter a link there)?

I have been struggling for years with Gmail's habit of inserting links into
my incoming mail.  In a past project, I had to analyse data sent in by mail
as rows of numbers; Gmail insists on turning some of them into links to (non
existent) phone numbers and addresses, which greatly complicates automatic
analysis.  (I'd love to hear from anyone who knows how to turn this off).

Please report problems with the web pages to the maintainer