What Facebook, Google and Twitter Told Congress About Russian Misinformation
January 25 2018 - 10:09PM
Dow Jones News
A Wall Street Journal Roundup
Congress on Thursday published responses from Facebook Inc.,
Twitter Inc. and Google to questions about how Russian actors used
their platforms to spread misinformation before and after the 2016
U.S. presidential election.
The responses address issues including whether there is any
evidence of collusion between the Russian parties and the Trump
campaign, and how Google, a unit of Alphabet Inc., is handling its
commercial transactions with a Russian broadcaster that federal
intelligence agencies say is a propaganda outlet for the
Kremlin.
The responses, which were submitted to Congress in recent weeks,
are an extension of answers that executives from the three
companies gave in hours of testimony last year. Here are
highlights, including which important questions the companies
didn't answer:
Facebook:
--The Internet Research Agency, a so-called Russian troll farm,
used Facebook's tools to promote rallies, protests and other events
across the U.S. that cut across ideological lines. In the responses
published Thursday, Facebook said 13 of the pages created by IRA
attempted to organize 129 events, about twice what The Wall Street
Journal tracked down this fall through archived versions of pages.
The events were viewed by 338,300 unique Facebook accounts, the
company said. Facebook said about 62,500 marked they were attending
one of the events and 25,800 accounts marked they were
interested.
--Facebook gave different responses to questions from two
lawmakers about whether there was evidence of collusion between the
Trump campaign and Russian-backed actors. In one set of answers,
the company said it found "what appears to be insignificant
overlap" between the ad-targeting criteria and content used by the
IRA and that used by the Trump campaign. In another set of
responses, Facebook said it can't "substantiate or disprove
allegations of possible collusion." Facebook also said that it had
no evidence that the IRA used U.S. voter registration data to
target ads on the platform.
--Last year, the Journal reported that the Russian government
was using software made by Moscow-based company Kaspersky Lab to
scan computers around the world. In its responses, Facebook said it
stopped using one Kaspersky antivirus products in October 2017 and
is in the process of "phasing out" another antivirus product used
internally that doesn't transmit data back to Kaspersky. Facebook
intends to continue use of a "Kaspersky service that provides us
with information about threat activity as a one-way feed."
-- Sen. Kamala Harris (D., Calif.) asked Facebook if it had seen
any evidence of state-sponsored manipulation in U.S. elections last
year, including gubernatorial races New Jersey and Virginia.
Facebook didn't directly answer the question, instead detailing how
it has learned and changed since the 2016 election cycle.
--Deepa Seetharaman
Twitter:
--Twitter gave its justification for why researchers estimate
10% to 15% of accounts on Twitter are spam, while Twitter itself
says fewer than 5% of accounts on its platform are spam. Twitter
said its own researchers have access to data about its users, such
as email addresses and phone numbers, that outside researchers
don't get -- and that this data helps Twitter tally spam accounts.
Twitter also said outside researchers sometimes count accounts as
spam that Twitter removes.
--Twitter defended its practice of not alerting users when they
are interacting with bot accounts. A label for bot accounts would
be too broad, Twitter said, because it might inadvertently include
nonmalicious activity, such as scheduled Twitter posts by humans,
and emergency notifications. Investigators, however, didn't say
such a label would necessarily indicate content is malicious.
--Twitter said a partnership among tech companies called the
Global Internet Forum to Counter Terrorism now includes a database
of 40,000 digitized fingerprints used to identify violent terrorist
imagery and recruitment videos. The group started with Twitter,
YouTube, Facebook and Microsoft Corp., and now also includes Snap
Inc., Microsoft's LinkedIn and others.
--Georgia Wells
Google:
--RT, the Russian state news agency that federal intelligence
officials have called "the Kremlin's principal international
propaganda outlet," uses Google's YouTube, Facebook and Twitter as
primary distributors of its content. Google told the Senate that it
now describes RT's relationship with the Russian government in
search results and that it is "working on disclosures to provide
similar transparency on YouTube." A Google search result for "RT"
on Thursday noted it is funded by the Russian government, but
search results that linked to RT articles didn't include such a
disclosure.
--One issue that is still unclear is how much Google and RT have
earned together on ads running before RT's YouTube videos begin
playing. As of late last year, RT had amassed 5.5 billion views on
YouTube and was part of the site's package of premium channels for
advertisers. YouTube has since removed RT from that package, but
still runs ads on its videos. Sen. Harris asked Google how much
revenue it earned on ads "that accompanied Russian propaganda."
Google responded it paid less than $35 to Russian actors who posted
videos on YouTube -- but it didn't disclose revenue for RT.
--Google said Russian actors that bought $4,700 worth of ads
around the election didn't "narrowly" target those ads to specific
groups of users, including by geography or political
affiliation.
--Google said in its responses that it hadn't detected
installation of Kaspersky products on its systems.
--Jack Nicas
(END) Dow Jones Newswires
January 25, 2018 21:54 ET (02:54 GMT)
Copyright (c) 2018 Dow Jones & Company, Inc.
Meta Platforms (NASDAQ:META)
Historical Stock Chart
From Mar 2024 to Apr 2024
Meta Platforms (NASDAQ:META)
Historical Stock Chart
From Apr 2023 to Apr 2024