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 01

Wednesday 14 December 2016


More on the LaMia crash involving the Brazilian soccer team
PwC SAP fatal flaw in security software
Iain Thomson via Al Mac
Netgear R7000 and R6400 vulnerability
Bob Gezelter
Automated Assistants Will Soon Make a Bid for Your Finances
Nathaniel Popper
Cars Talking to One Another? They Could Under Proposed Safety Rules
Cecilia Kang
ACLU sues Rhode Island over computer benefits system delays
AP item via The Boston Globe
Designing a Safer Battery for Smartphones—That Won't Catch Fire
John Markoff
Fake News Expert On How False Stories Spread And Why People Believe Them
SHAME ON YOU, GOOGLE! - Holocaust Deniers as the top search result
Gizmodo via Lauren Weinstein
Europe braces for Russian hacking in upcoming elections
Russia hacking the DNC
The New York Times
On the CIA assessment: Russia intervened in the 2016 election
Peter Houppermans
The Perfect Weapon: How Russian Cyberpower Invaded the U.S.
The New York Times
Don't like a political blog? Go after their advertising revenue
Thomas Koenig
Trump's F-35 tweet sends Lockheed Martin stock into tailspin
Steve Bittenbender
Ashley Madison settles cheaply for $1.6 million
Re: Boeing Dreamliner 787 should be reboot every 21 days
Michael Kohne
Re: Ball-bearing and crypto policy analogy
Serguei Patchkovskii
Ron Rivest
Info on RISKS (comp.risks)

More on the LaMia crash involving the Brazilian soccer team

"Alister Wm Macintyre \(Wow\)" <>
Tue, 13 Dec 2016 02:29:18 -0600
A plane crash, killing almost an entire Brazilian football team, has been
explained.  The plane operators violated some standards.  They neglected to
have a refueling stop, and the plane plain ran out of fuel There's been some
finger pointing about that.  An airport official said she warned the plane
crew that they needed to fuel up before leaving, but the crew assured her
they had enough.  Gov blaming her for not doing what she said she did, so
she has fled across a border seeking asylum.

PwC SAP fatal flaw in security software (Iain Thomson)

"Alister Wm Macintyre \(Wow\)" <>
Sat, 10 Dec 2016 22:07:07 -0600
Iain Thomson, *The Register*, 9 Dec 2016

PwC has issued a denial that there is anything wrong with their software.
How do we know there's any truth in their denial?

I suppose it is inconceivable to an audit firm that anyone ought to audit

Normally when flaws are found in a corporate software package, clients
report the problem to tech support, and the situation gets fixed, and the
fix can be tested.

Here a company is not providing normal industry standard support.  They want
people to take their word for it that their software is fine, even when
evidence has been revealed to them that there is a problem.  This is
reminiscent of the Volkswagen cover-up that their cars could be stolen via
hacking the auto door locks.  Did they ever fix that?

  Iain Thomson, *The Register*, 9 Dec 2016
  Fatal flaw found in PricewaterhouseCoopers SAP security software
  Instead of fixing the issue, PwC lawyered up

A security tool built for SAP systems by PricewaterhouseCoopers has turned
out to have worrying security holes of its own.

German security research firm ESNC has been analyzing the Automated Controls
Evaluator (ACE), which extracts relevant security and configuration data
from an SAP system, analyzes it, and generates exception reports by review.
But there appears to be a high-risk hole in the software.

"This security vulnerability may allow an attacker to manipulate accounting
documents and financial results, bypass change management controls, and
bypass segregation of duties restrictions," ESNC said in an advisory.>

"This activity may result in fraud, theft or manipulation of sensitive data
including PII such as customer master data and HR payroll information,
unauthorized payment transactions and transfer of money."

Comments to the Register article ask:

* How the PWC software can be so badly written as to allow this to happen?
  Does it have anything to do with the company being run by non-tech people?

* How PWC can be so clueless about fixing flawed software, that they'd
  rather lawyer up than fix it?  ESNC gave them 90 days after discovery and
  notification, before going public.

* The next time anyone finds a PWC vulnerability, they won't do them the
  courtesy of notification & reasonable time to fix, they'll just go public
  to warn other PWC customers.

* Search for "PWC scandal" to find lots of times this company has been in
  big trouble already.

* There was a question about lawyer hacker vulnerability.  Someone who must
  be unaware that there has already been massive hacking of major law firms,
  to facilitate such things as crooked insider trading, and telling the
  world about Panama Papers.

Here's info about SAP:

For a company to be vulnerable to this breach vulnerability, they'd have to
be running on SAP with the PWC's ACE

Here's directory of industries served by PWC:

Netgear R7000 and R6400 vulnerability

"Bob Gezelter" <>
Mon, 12 Dec 2016 02:46:01 -0700
Another installment from the "When will they ever learn" files:

Netgear R7000 and R6400 routers have been found to contain an "arbitrary
command injection" vulnerability.

CERT Vulnerability Note VU#582384, entitled "Multiple Netgear routers are
vulnerable to arbitrary command injection" describes the details of the the
vulnerability, for which an exploit example is available.

As reported by the CERT notice, there is presently no corrected firmware
available for the devices. CERT recommends that the use of affected devices
be discontinued until such time as a fix is available.

The CERT Notice can be found at:

Bob Gezelter,

Automated Assistants Will Soon Make a Bid for Your Finances (Nathaniel Popper)

Monty Solomon <>
Wed, 14 Dec 2016 09:57:09 -0500
Nathaniel Popper, The New York Times, 7 Dec 2016

Companies are vying to create automated financial assistants that employ
artificial intelligence; one was directly inspired by science fiction.

Cars Talking to One Another? They Could Under Proposed Safety Rules (Cecilia Kang)

Monty Solomon <>
Tue, 13 Dec 2016 21:29:59 -0500
Cecilia Kang, The New York Times, 13 Dec 2016

Under the rules, cars would be able to use wireless technology to detect if
another vehicle was moving too fast in their direction and headed for a

ACLU sues Rhode Island over computer benefits system delays (AP)

Monty Solomon <>
Sun, 11 Dec 2016 12:36:13 -0500
AP item via The Boston Globe, 9 Dec 2016

Designing a Safer Battery for Smartphones—That Won't Catch Fire

Monty Solomon <>
Sun, 11 Dec 2016 23:09:44 -0500
John Markoff, *The New York Times*, 11 Dec 2016
A Massachusetts start-up is part of a new wave of efforts in the United
States, Europe, and Asia to improve battery technologies as consumers demand
more from phones and cars.

Fake News Expert On How False Stories Spread And Why People Believe Them (NPR)

Lauren Weinstein <>
Wed, 14 Dec 2016 12:28:58 -0800
via NNSquad

  Craig Silverman of BuzzFeed News has spent years studying media
  inaccuracy. He explains how false stories during the presidential campaign
  were spread on Facebook and monetized by Google AdSense.

SHAME ON YOU, GOOGLE! - Holocaust Deniers as the top search result

Lauren Weinstein <>
Mon, 12 Dec 2016 21:38:49 -0800
Google Won't Alter the Holocaust-Denying Results For 'Did the Holocaust

SHAME ON YOU, GOOGLE! - While I agree with your decision to not remove the
lying hate speech link in question, you should clearly label it as being
false, a lie, or at least as having no credibility.  Call it "CredRank" Zero
if you wish, but the fact is that most users of Google implicitly trust you
so much that they assume you wouldn't rank vile, lying crap at the top of
your search results.

You know and I know that those top results don't mean that they are
"correct"—and they don't mean that you endorse them.  But it is widely
believed that what Google puts at the top can be trusted. Once upon a time,
you dealt with the search term "Jew" by including a note about related hate

The time has come for Google to lead the way against hate speech and fake
news. Here's how I hope you will do so: "Action Items: What Google,
Facebook, and Others Should Be Doing RIGHT NOW About Fake News":

  See also:

Europe braces for Russian hacking in upcoming elections

"Peter G. Neumann" <>
Tue, 13 Dec 2016 7:34:15 PST
Officials fear cyber-meddling by Moscow in upcoming elections in France,
the Netherlands and Germany.

Politico's cybersecurity newsletter today + an alternative intelligence view
re direct Russian involvement

COMMISSIONS, SELECT COMMITTEES AND MORE - There are now no fewer than five
different proposals for how Congress might push an investigation into
alleged Russian election meddling and related cybersecurity issues.
Sens. Ben Cardin, Dianne Feinstein and Patrick Leahy on Monday proposed an
independent commission, with a different name but similar makeup to one
proposed in the House by Reps. Eric Swalwell and Elijah Cummings. Sen. Cory
Gardner on Monday again called for the creation of a Permanent Select
Committee on Cybersecurity, inspired in part by the campaign hacks.  Senate
Armed Services Chairman John McCain over the weekend suggested a select
committee that would exist only temporarily to investigate election hacking.

Some of those proposals might yet become reality, but what looks most likely
in the near term is the idea endorsed by Senate Majority Leader Mitch
McConnell, where the Senate Intelligence Committee would lead an
investigation into potential foreign influence in the election and Senate
Armed Services delving into the more general threat of cyberattacks.

In the House, the most likely result is no special investigation at all. [...]

Russia hacking the DNC

"Peter G. Neumann" <>
Tue, 13 Dec 2016 14:00:27 PST

Hundreds of similar phishing emails were being sent to American political
targets, including an identical email sent on March 19 to Mr. Podesta,
chairman of the Clinton campaign.  Given how many emails Mr. Podesta
received through this personal email account, several aides also had access
to it, and one of them noticed the warning email, sending it to a computer
technician to make sure it was legitimate before anyone clicked on the
"change password" button.

"This is a legitimate email," Charles Delavan, a Clinton campaign aide,
replied to another of Mr. Podesta's aides, who had noticed the alert.  "John
needs to change his password immediately."

With another click, a decade of emails that Mr. Podesta maintained in his
Gmail account—a total of about 60,000 - were unlocked for the Russian
hackers. Mr. Delavan, in an interview, said that his bad advice was a result
of a typo: He knew this was a phishing attack, as the campaign was getting
dozens of them. He said he had meant to type that it was an "illegitimate"
email, an error that he said has plagued him ever since.

On the CIA assessment: Russia intervened in the 2016 election (R 29 96)

Peter Houppermans <>
Sun, 11 Dec 2016 16:11:22 +0100
Pardon me for maybe missing something, but is Russia's (possibly) hacking
the election really the key problem?

The issue is not that Russia has (possibly) hacked the election, the issue
is that it is deemed perfectly possible it could.

I may be kicking in an open door here, but if a vital democratic mechanism
is so mistrusted that any statement of it being hacked is deemed credible
(and from the reports I've seen of some voting systems there's indeed reason
to believe it possible), isn't that a big hint that things need fixing
rather urgently?

Writing accusingly about an increase of burglaries in your neighbourhood
might sell more newspapers but personally, I would rather make sure my locks
are up to scratch.

  [Many locks are vulnerable, and they should be scratched!  PGN]

The Perfect Weapon: How Russian Cyberpower Invaded the U.S.

Monty Solomon <>
Tue, 13 Dec 2016 22:06:25 -0500
Eric Lipton, David E. Sanger and Scott Shane, *The New York Times*, 13 Dec 2016

An investigation by *The New York Times* reveals missed signals, slow
responses and a continuing underestimation of the seriousness of a campaign
to disrupt the 2016 presidential election.

Don't like a political blog? Go after their advertising revenue

Thomas Koenig <>
Mon, 12 Dec 2016 23:18:32 +0100
In Germany, there is an Internet campaign to bring down political blogs
considered to be "right-wing"; its hashtag is #KeinGeldfuerRechts (no money
for the right wing).

The campaign contacts companies whose advertising is displayed on these
websites, and ask them to consider if they really want their names to be
displayed on these websites.

Some of the blogs that have seen advertising revenues drop dramatically due
to this campaign are "Die Achse des Guten" (the Axis of Good, and "Tichys Einblick" (Tichy's insight,

The campaign is headed by an advertising executive, Gerald Hensel, who works
for Scholz & Partners. The company is currently suffering something of a
sh..storm for failing to distance itself sufficiently from their executive.
In the meantime, the website calling for the advertising boycott,, has been switched to password-only access.

The risks?  Trying to shut up your political opposition by targeting
their advertising funds may work (which is not a pleasant thought), or
it may backfire.

Trump's F-35 tweet sends Lockheed Martin stock into tailspin (Steve Bittenbender)

"Peter G. Neumann" <>
Tue, 13 Dec 2016 8:10:59 PST
Steve Bittenbender, Government Security News, 13 Dec 2016

On the same day Lockheed Martin delivered a two F-35s to Israel,
President-elect Donald Trump took the country's largest government
contractor to task for its handling of the fighter jet program's finances.

The F-35 program and cost is out of control.  Billions of dollars can and
will be saved on military (and other) purchases after January 20th," Trump
posted on Twitter Monday morning. [...]

Ashley Madison settles cheaply for $1.6 million (FTC)

"Peter G. Neumann" <>
Wed, 14 Dec 2016 13:12:17 PST
   (Previous item in RISKS-29.63:
     Ashley Madison Admits It Lured Customers With 70,000 Fake 'Fembots'

Federal Trade Commission

The operators of the Toronto-based dating site have agreed
to settle Federal Trade Commission and state charges that they deceived
consumers and failed to protect 36 million users' account and profile
information in relation to a massive July 2015 data breach of their
network.  The site has members from over 46 countries.

The settlement requires the defendants to implement a comprehensive
data-security program, including third-party assessments.  In addition, the
operators will pay a total of $1.6 million to settle FTC and state actions.

"This case represents one of the largest data breaches that the FTC has
investigated to date, implicating 36 million individuals worldwide," said
FTC Chairwoman Edith Ramirez.  "The global settlement requires to implement a range of more robust data security
practices that will better-protect its users' personal information from
criminal hackers going forward."

In addition to the provisions prohibiting the alleged misrepresentations and
requiring a comprehensive security program, the proposed federal court order
imposes an $8.75 million judgment which will be partially suspended upon
payment of $828,500 to the Commission.  If the defendants are later found to
have misrepresented their financial condition, the full amount will
immediately become due.  An additional $828,500 will be paid to the 13 states
and the District of Columbia.

Re: Boeing Dreamliner 787 should be reboot every 21 days (PGN, RISKS-29.96)

Michael Kohne <>
Wed, 14 Dec 2016 11:30:48 -0500
I have a couple of thoughts on why it might not be fixed yet. I've never
done software for aircraft, just for medical devices (so my software has
never been able to kill more than one person at a time):

1) I don't know what the lead time on a software release for an aircraft
   is. I'm betting their review and testing rules are pretty tight and take
   quite a while. Even if they've got the bug fixed, it may take quite some
   time to see the fix in the field.

2) We don't know what, exactly, is going on, but assuming it's the signed
   value as described, it seems likely that it could take quite a while to
   be sure you've got all the instances where those time values are
   mis-used. Depending on how use of that value is structured (for instance,
   the routine that returns time might be returning a signed value), fixing
   it might end up touching large portions of the system, thereby triggering
   massive amounts of code review.

3) Even if they fix it, are they sure enough of the fix? I'm sure it's
   tempting for Boeing to say 'well, we'll roll out the fix, but keep the
   reboot rule so that if we missed anything we don't get blamed'.

4) Even if there's a fix, the airlines may not have rolled it out.  I've no
   idea what an airline does for software patching a plane, but I'm betting
   it's a more complex endeavor than just getting the files from Boeing and
   taking them out to the plane.

So there's a lot of reasons why a fix might not be in the field yet.

Re: Ball-bearing and crypto policy analogy (Rivest, RISKS-29.96)

Serguei Patchkovskii <>
Sun, 11 Dec 2016 09:21:06 +0100
Ronald Rivest has suggested an interesting analogy between law-enforcement
agencies controlling cryptographic techniques and similar controls being
imposed on ball bearings.

I think this analogy is actually much closer than intended: The specific
examples given in the item make ball-bearing controls sound completely
nonsensical.  However, high-grade ball bearings and related manufacturing
equipment *are* in fact quite tightly controlled, and with some good
reasons.  The US Department of Commerce list of export controls on ball
bearings and related technologies runs to some ten pages:]
Similar restrictions are imposed by all countries participating in Wassenaar
Violating these rules can land you in some serious trouble.

Re: Ball-bearing and crypto policy analogy (Patchkovskii, RISKS-30.01)

"Ronald L. Rivest" <>
Mon, 12 Dec 2016 15:57:33 -0800
Thanks to Serguei Patchkovskii for the information regarding the controls on
the export of ball bearings.  I was unaware of the existence of these

The controls on ball bearings have to do with their tolerances primarily.
The cryptographic analogue would probably be a control on key-size.  Since
ball bearings are to be part of a manufactured product, while cryptographic
schemes are there to defeat and adversarial attack, the restriction of
commercial users to 'weak' crypto isn't really a good idea.

Please report problems with the web pages to the maintainer