Hearings
Hearing Type:
Open
Date & Time:
Wednesday, November 1, 2017 - 9:30am
Location:
Hart 216
Witnesses
Vice President and General Counsel
Colin
Stretch
Facebook
General Counsel
Sean
Edgett
Twitter
Senior Vice President and General Counsel
Kent
Walker
Google
Social Media Influence in the 2016 U.S. Elections Exhibits
Full Transcript
[Senate Hearing 115-232]
[From the U.S. Government Publishing Office]
S. Hrg. 115-232
OPEN HEARING: SOCIAL MEDIA INFLUENCE
IN THE 2016 U.S. ELECTION
=======================================================================
HEARING
BEFORE THE
SELECT COMMITTEE ON INTELLIGENCE
OF THE
UNITED STATES SENATE
ONE HUNDRED FIFTEENTH CONGRESS
FIRST SESSION
__________
WEDNESDAY, NOVEMBER 1, 2017
__________
[GRAPHIC NOT AVAILABLE IN TIFF FORMAT]
Printed for the use of the Select Committee on Intelligence
Available via the World Wide Web: http://www.govinfo.gov
U.S. GOVERNMENT PUBLISHING OFFICE
27-398 PDF WASHINGTON : 2018
----------------------------------------------------------------------------------------
For sale by the Superintendent of Documents, U.S. Government Publishing Office,
http://bookstore.gpo.gov. For more information, contact the GPO Customer Contact Center,
U.S. Government Publishing Office. Phone 202-512-1800, or 866-512-1800 (toll-free).
E-mail, gpo@custhelp.com.
SELECT COMMITTEE ON INTELLIGENCE
[Established by S. Res. 400, 94th Cong., 2d Sess.]
RICHARD BURR, North Carolina, Chairman
MARK R. WARNER, Virginia, Vice Chairman
JAMES E. RISCH, Idaho DIANNE FEINSTEIN, California
MARCO RUBIO, Florida RON WYDEN, Oregon
SUSAN COLLINS, Maine MARTIN HEINRICH, New Mexico
ROY BLUNT, Missouri ANGUS KING, Maine
JAMES LANKFORD, Oklahoma JOE MANCHIN III, West Virginia
TOM COTTON, Arkansas KAMALA HARRIS, California
JOHN CORNYN, Texas
MITCH McCONNELL, Kentucky, Ex Officio
CHUCK SCHUMER, New York, Ex Officio
JOHN McCAIN, Arizona, Ex Officio
JACK REED, Rhode Island, Ex Officio
----------
Chris Joyner, Staff Director
Michael Casey, Minority Staff Director
Kelsey Stroud Bailey, Chief Clerk
CONTENTS
----------
NOVEMBER 1, 2017
OPENING STATEMENTS
Burr, Hon. Richard, Chairman, a U.S. Senator from North Carolina. 1
Warner, Hon. Mark R., Vice Chairman, a U.S. Senator from Virginia 4
WITNESSES
Stretch, Colin, Vice President and General Counsel, Facebook..... 7
Prepared statement........................................... 9
Edgett, Sean, General Counsel, Twitter........................... 16
Prepared statement........................................... 18
Walker, Kent, Senior Vice President and General Counsel, Google.. 38
Prepared statement............................................... 41
SUPPLEMENTAL MATERIAL
Exhibits used by Chairman Burr................................... 47
Exhibits used by Vice Chairman Warner............................ 54
Exhibits used by Senator Collins................................. 69
Exhibits used by Senator King.................................... 79
Answers to questions for the record from Colin Stretch........... 100
Answers to questions for the record from Sean Edgett............. 130
Answers to questions for the record from Kent Walker............. 173
Submission from Senator Harris................................... 199
OPEN HEARING: SOCIAL MEDIA INFLUENCE IN THE 2016 U.S. ELECTION
----------
WEDNESDAY, NOVEMBER 1, 2017
U.S. Senate,
Select Committee on Intelligence,
Washington, DC.
The Committee met, pursuant to notice, at 9:34 a.m. in Room
SH-216, Hart Senate Office Building, Hon. Richard Burr
(Chairman of the Committee) presiding.
Committee Members Present: Senators Burr, Warner, Risch,
Rubio, Collins, Blunt, Lankford, Cotton, Cornyn, Feinstein,
Wyden, Heinrich, King, Manchin, Harris, and Reed.
OPENING STATEMENT OF HON. RICHARD BURR, CHAIRMAN, A U.S.
SENATOR FROM NORTH CAROLINA
Chairman Burr. I'd like to call the hearing to order. Good
morning.
I'd like to welcome our witnesses today. Before I introduce
them, I want to say, on behalf of the full Committee, that our
hearts and our prayers go out to the individuals in New York,
the families and the friends of those who were affected by a
senseless terror act. To most on this Committee, we've come to
expect this. We spend countless hours working through the
threats that exist to this country and around the world, and
it's sad that we've come to the point where, really, nothing
can happen that surprises us.
But it's the responsibility of this Committee to work hand-
in-hand with our intelligence community to help to keep America
safe by providing the tools that they need to accomplish their
mission. We will continue to do that.
As is the case that we're here today, and I welcome our
witnesses, Colin Stretch, Vice President and General Counsel at
Facebook; Sean Edgett, General Counsel at Twitter; and Kent
Walker, Senior Vice President, General Counsel at Google.
For several months now, the media has been fixated on the
role that social media platforms played in spreading
disinformation and discord during the 2016 elections. This is
an opportunity for each of you to tell your respective stories
and, if necessary, correct the record. My sense is that not all
aspects of those stories have been told accurately. I'll note
for the record that this Committee is now having its
seventeenth open hearing this year, and the twelfth at which
we'll be discussing Russia and Russia's activities.
Today, I'm hopeful we can provide the American people with
an informed and credible assessment of how foreign actors used
your platforms to circulate lies and to agitate unrest during
last year's elections.
I'm also hopeful you'll share with us what your companies
are doing to make it harder for foreign actors to use your
platforms' automated accounts and falsified stories to
influence sentiment in the United States.
Very clearly, this kind of national security vulnerability
represents an unacceptable risk, and your companies have a
responsibility to reduce that vulnerability. While we're on the
topic of responsibility, I want to use this forum to push back
on some narratives that have sprung up around the subject. A
lot of folks, including many in the media, have tried to reduce
this entire conversation to one premise; foreign actors
conducted a surgical, executed covert operation to help elect a
United States president.
I'm here to tell you this story does not simplify that
easily. It is shortsighted and dangerous to selectively focus
on one piece of information and think that that somehow tells
the whole story.
We've heard from the media how a series of, I quote,
``Russian-linked Facebook ads were specifically aimed at
Michigan and Wisconsin during the lead-up to last year's
presidential election,'' unquote, and that, quote, ``some of
those ads targeted specific demographic groups in two states,''
unquote. The narrative here is that ads linked to Russia were
targeted at pivotal states and directly influenced the
election's outcome.
What you haven't heard is that almost five times more ads
were targeted at the State of Maryland than of Wisconsin,
Maryland, which is targeted by 262 ads in comparison to
Wisconsin's 55 ads, and Maryland was not up for grabs. It was a
State the Democrat candidate carried by 26 percent; or that 35
of the 55 ads targeted at Wisconsin ran prior to the Wisconsin
primary, before there was an identified Republican candidate;
and moreover, that not one of those 55 ads mentioned President
Donald Trump by name; or that the key election State of
Pennsylvania had fewer ads targeted at it than Washington,
D.C., where 87 percent of the electorate voted for Hillary
Clinton; or that the three most heavily targeted states in
America--Maryland, Missouri, and New York--were all determined
by at least 18-point margin, and two of them won by Hillary
Clinton.
One point the media has gotten correct is that more of
these geographically targeted ads ran in 2015 than 2016--again,
before President Trump was identified as the Republican
candidate for president. But some of the context surrounding
the more than $100,000 worth of divisive ads on hot-button
issues purchased by Russian actors is missing. To add some
detail here where the media has failed to do it and put the
$100,000 into a frame of reference, the total ad spend for the
State of Wisconsin was $1,979, with all but $54 being spent
before the primary--again, before the emergence of a Republican
candidate. The ad spend in the State of Michigan was $823;
Pennsylvania, $300.
To believe the narrative, you have to accept that these
sophisticated, well-resourced Russian actors studied our
process, assessed what states would be critical to the election
result, then snuck and invested all of $300 to execute their
plan in Pennsylvania--$300. More than five times as much money
was spent on advertising in California, a State that hasn't
voted Republican in presidential elections since 1988.
Even with the benefit of numbers and what can be calculated
and measured, this is an incredibly complex story. We can look
at the amount of money spent, the number of ads purchased, and
draw conclusions about priorities. We can look at the divisive
content of the ads and the pages that they directed people
towards, the number of tweets and retweets, and the manipulated
search results, and draw inferences about the intent of the
information operation. What we cannot do, however, is calculate
the impact that foreign meddling in social media had on this
election, nor can we assume that it must be the explanation for
an election outcome that many didn't expect.
I understand the urge to make this story simple. It's human
nature to make the complex manageable, find explanations, and
interpret things in ways that conform to your conclusions. But
that's biased. Pointing to a State and saying that no ads ran
there after the election doesn't prove intent, or even motive.
It just shows that no ads ran there after the election.
This subject is complicated. There's a whole new vocabulary
that comes with this stuff. Impressions are different than
views. Views are different than clicks. But there's one thing
that I'm certain of and it's this: Given the complexity of what
we've seen, if anyone tells you they've got this all figured
out, they're kidding themselves. And we can't afford to kid
ourselves about what happened last year and continues to happen
today.
That complexity, I'll note, is exactly why we depend on you
for expert insight and reliable information. Sixty percent of
the U.S. population uses Facebook. A foreign power using that
platform to influence how Americans see and think about one
another is as much a public policy issue as it is a national
security concern. Crafting an elegant policy solution that is
effective, but not overly burdensome, demands good faith and
partnership between companies and this Committee.
Just recently, on the basis of a more complete and
sophisticated analysis, the original estimate that 10 million
Americans were exposed to Russian-origin content on Facebook
was increased to 126 million. That tells me that your companies
are just beginning to come to grips with the scale and the
depth of the problem.
That's encouraging, but know this: we do better when you do
better. I'd urge you to keep that in mind and to work with us
proactively to find the right solution to a very constant and
complaining challenge.
I'll take a moment here to stress what this hearing is and
is not about. This isn't about relitigating the 2016 U.S.
presidential election. This isn't about who won or who lost.
This is about national security. This is about corporate
responsibility, and this is about the deliberative and
multifaceted manipulation of the American people by agents of a
hostile foreign power.
I'll say it again: agents of a hostile foreign power
reached into the United States, using our own social media
platforms, and conducted an information operation intended to
divide our society along issues like race, immigration and
Second Amendment rights. What's even more galling is that, to
tear us apart, they're using social media platforms Americans
invented, in connection with the First Amendment freedoms that
define an open and democratic society.
While it's shocking to think that foreign actors use the
social networking and communications mediums that are so
central to our lives today in an effort to interfere with the
core of our democracy, what is even more troubling is the
likelihood that these platforms are still being used today to
spread lies, provoke conflict and drive Americans apart.
Your three companies have developed platforms that have
tremendous reach and, therefore, tremendous influence. That
reach and influence is enabled by the enormous amount of data
you collect on your users and their activities. The American
people now need to understand how Russia used that information
and what you're doing to protect them. Your actions need to
catch up to your responsibilities.
We have a lot to get to this morning. I'm going to stop
here. Again, I want to thank each of our briefers--our
witnesses today, and I turn to the vice chairman for any
comments he might have.
OPENING STATEMENT OF HON. MARK R. WARNER, A U.S. SENATOR FROM
VIRGINIA
Vice Chairman Warner. Well, thank you, Mr. Chairman, and
let me also express our concerned thoughts about the tragedy
yesterday in New York.
Let me get right at it. In the age of social media, you
can't afford to waste too much time or, for that matter, too
many characters, in getting the point across. So I'll get
straight to the bottom line: Russian operatives are attempting
to infiltrate and manipulate American social media to hijack
the national conversation and to make Americans angry; to set
us against ourselves, and, at their most basic, to undermine
our democracy. They did it during the 2016 U.S. presidential
campaign. They are still doing it now. And not one of us is
doing enough to stop it. That's why we're here today.
In many ways, the threat is not new. Russians have been
conducting information warfare for decades. But what is new is
the advent of social media tools with the power to magnify
propaganda and fake news on a scale that was unimaginable back
in the days of the Berlin Wall.
Today's tools in many ways seem almost purpose-built for
Russian disinformation techniques. Russia's playbook is simple,
but formidable. It works like this.
First, disinformation agents set up thousands of fake
accounts, groups and pages across a wide array of platforms.
These fake accounts populate content on Facebook, Instagram,
Twitter, YouTube, Reddit, LinkedIn, and many other platforms.
Each of these fake accounts spends literally months
developing networks of real people to follow and like their
content, boosted by tools like paid ads and automated bots.
Most of the real-life followers have no idea that they are
caught up in these webs. These networks are later utilized to
push an array of disinformation, including stolen e-mails,
state-led propaganda like RT News and Sputnik, fake news, and
divisive content.
The goal is pretty simple. It's to get this so-called news
into the news feeds of many potentially receptive Americans and
to covertly and subtly push those Americans in the directions
the Kremlin wants to go.
As someone who deeply respects the tech industry and who
was involved in that industry for more than 20 years, it's
taken me quite a bit of time--and I'm still learning--to truly
understand the nature of this threat. Even I struggle to keep
up with the language and the mechanics, the difference between
bots, trolls, and fake accounts; how they generate likes,
tweets, and shares; and how all these players and actions are
combined into an online ecosystem.
What is clear, however, is that this playbook offers a
tremendous bang for the disinformation buck. With just a small
amount of money, adversaries use hackers to steal and weaponize
data, trolls to craft disinformation, fake accounts to build
networks, bots to drive traffic, and ads to target new
audiences. They can force propaganda into the mainstream and
wreak havoc on our online discourse. And if you look back at
the results, it's a pretty good return on investment.
So where do we go from here? I believe it will take all of
us--you, some of the platform companies, the United States
government, and the American people--to deal with this new and
evolving threat.
The social media and innovative tools each of you have
developed have changed our world for the better. You've
transformed the way we do everything from shopping for
groceries to growing small businesses.
But Russia's actions are further exposing the dark
underbelly of the ecosystem you have created, and there is no
doubt that their successful campaign will be replicated by
other adversaries--both nation-states and terrorists--that wish
to do harm to democracies around the globe. This is not a
unique American phenomenon.
As such, each of you here today needs to commit more
resources to identifying bad actors and, when possible,
preventing them from abusing our social media ecosystem. Thanks
in part to pressure from this Committee, each company has
uncovered, I believe, only some of the evidence of the ways
Russians exploited their platforms during the 2016 election.
For Facebook, much of the attention has been focused on the
paid ads that Russian trolls targeted to Americans. However,
these ads are just the tip of a very large iceberg. The real
story is the amount of misinformation and divisive content that
was pushed for free on Russian-backed pages, which was then
spread widely on news feeds of tens of millions of Americans.
According to the data Facebook has provided, 120 Russian-
backed pages built a network of over 3.3 million people. From
these now-suspended pages, 80,000 organic unpaid posts reached
an estimated 126 million real people--more than a third of the
population.
This is an astonishing reach from just one group in St.
Petersburg. And I doubt that the so-called Internet Research
Agency in St. Petersburg represents the only Russian trolls out
there.
Facebook has more work to do to see how deep this goes,
including into the reach that we've just found in the last 48
hours of information you've provided, of IRA-backed Instagram
posts, which, again, if we just take for an example, 80,000
posts from IRA-based trolls on Facebook, 120,000 pieces of
content on Instagram, and we don't even have the data on how
many--how much that content reached.
The anonymity provided by Twitter and the speed by which it
shares news makes it an ideal tool to spread disinformation.
According to one study, during the 2016 campaign, junk news
actually outperformed real news in some battleground states,
leading up to Election Day. Another study found that bots
generated one out of every five political messages posted on
Twitter over the entire presidential campaign.
I'm concerned, sir, that Twitter seems to be vastly
underestimating the number of fake accounts and bots pushing
disinformation. Independent researchers, people who've
testified before this Committee, have estimated that up to 15
percent of active Twitter accounts, or potentially 45 million-
plus accounts, are fake or automated.
Despite evidence of significant incursion and outreach from
researchers, Twitter has to date only uncovered a small piece
of that activity, although I will acknowledge that in the last
few days your numbers have gone from about 200 accounts to over
2,700 accounts. And again, I believe there's more to be done.
Google search algorithms continue to have problems in
surfacing fake news or propaganda. Though we can't necessarily
attribute to Russian efforts, false stories and unsubstantiated
rumors were elevated on Google Search during the recent mass
shootings in Las Vegas.
Meanwhile, YouTube has become RT's go-to platform. Google
has now uncovered 1,100 videos associated with this Russian
campaign. Much more of your content was likely spread through
other platforms.
But it's not just the platforms that need to do more. The
United States government has thus far proven incapable of
meeting this 21st-century challenge. Unfortunately, I believe
this effort is suffering in part because of lack of leadership
at the top. We have a President who remains unwilling to
acknowledge the threat that Russia poses to our democracy.
President Trump should stop actively delegitimizing American
journalism and acknowledge and address this very real threat
posed by Russian propaganda.
I believe that Congress, too, must do more. We need to
recognize that current law was not built to address these
threats. I partnered with Senators Klobuchar and McCain on what
I believe is the most light-touch legislative approach, which I
hope all my colleagues on this panel will review. The Honest
Ads Act is a national security bill intended to protect our
elections from the foreign interference we all want to avoid.
Finally, but perhaps most importantly, the American people
also need to be aware of what is happening to our news feeds.
We all need to take a more discerning approach to what we are
reading and sharing and who we're connecting with online. We
need to recognize the person at the other end of that Facebook
or Twitter argument may not be a real person at all.
The fact is that this Russian weapon has already proved its
success and cost-effectiveness. We can be assured that other
adversaries, including foreign intelligence operatives and
potentially terrorist organizations, have read this playbook
and are already taking action. It's why we, collectively, must
act.
To our witnesses today, I hope you will detail what we saw
in the last election and, most importantly, tell us what steps
you will undertake for us to get ready for the next one. We
welcome your participation and encourage your commitment to
addressing this shared responsibility.
Thank you, Mr. Chairman.
Chairman Burr. Thank, Senator Warner.
I'd like to notify Members we will have seven-minute rounds
today by seniority.
Gentlemen, if I could ask you to please stand and raise
your right hand. Do you solemnly swear to tell the truth, the
whole truth, and nothing but the truth?
Mr. Stretch. Yes.
Mr. Walker. I do.
Mr. Edgett. Yes.
Chairman Burr. Please be seated.
Mr. Stretch, we're going to recognize you, then Mr. Edgett,
then Mr. Walker. Mr. Stretch, the floor is yours.
STATEMENT OF COLIN STRETCH, VICE PRESIDENT AND GENERAL COUNSEL,
FACEBOOK
Mr. Stretch. Chairman Burr, Vice Chairman Warner and
distinguished Members of the Committee, thank you for this
opportunity to appear before you today. My name is Colin
Stretch and since July 2013 I've served as the General Counsel
of Facebook. We appreciate this Committee's hard work to
investigate Russian interference in the 2016 election.
At Facebook, our mission is to create technology that gives
people the power to build community and bring the world closer
together. We are proud that each of you uses Facebook to
connect with your constituents, and we understand that the
people you represent expect authentic experiences when they
come to our platform to share and connect.
We also believe that we have an important role to play in
the democratic process and a responsibility to protect it on
our platform. That's why we take what's happened on Facebook so
seriously.
The foreign interference we saw during the 2016 election is
reprehensible. That foreign actors hiding behind fake accounts
abused our platform and other internet services to try to sow
division and discord and to try to undermine our election
process is directly contrary to our values and everything we
stand for. Our goal at Facebook is to bring people closer
together. These foreign actors sought to drive people apart.
In our investigation, which continues to this day, we have
found that these actors used fake accounts to place ads on
Facebook and Instagram that reached millions of Americans over
a two-year period, and that those ads were used to promote
pages which in turn posted more content. People shared these
posts, spreading them still further.
Many of these ads and posts are inflammatory. Some are
downright offensive. We know that much of this content is
particularly hurtful to members of the Facebook community that
engaged with this content believing it was authentic. People
should believe content on Facebook is authentic, and should not
have to worry that they are being exploited in a cynical effort
to prey on painful fault lines in our society in order to
inflame discourse in this country.
In aggregate, the ads and posts we are here today to
discuss were a very small fraction of the overall content on
Facebook, but any amount is too much. All of these accounts and
pages violated our policies, and we removed them.
Going forward, we are making significant investments. We're
hiring more ad reviewers, doubling or more our security
engineering efforts, putting in place tighter ad content
restrictions, launching new tools to improve ad transparency,
and requiring documentation from political ad buyers.
We're building artificial intelligence to help locate more
banned content and bad actors. We're working more closely with
industry to share information on how to identify and prevent
threats, so that we can all respond faster and more
effectively. And we're expanding our efforts to work more
closely with law enforcement.
We know bad actors aren't going to stop their efforts. We
know we'll all have to keep learning and improving to stay
ahead of them. We also know we can't do this alone. That's why
I want to thank you for this investigation. We look forward to
the conclusions you will ultimately share with the American
public. And I look forward to your questions.
[The prepared statement of Mr. Stretch follows:]
[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]
Chairman Burr. Mr. Edgett, the floor is yours.
STATEMENT OF SEAN EDGETT, GENERAL COUNSEL, TWITTER
Mr. Edgett. Chairman Burr, Vice Chairman Warner and Members
of this Committee, Twitter understands the importance of the
Committee's inquiry into Russia's interference in the 2016
election, and we appreciate the opportunity to appear here
today. The events underlying this hearing have been deeply
concerning to our company and the broader Twitter community. We
are committed to providing a service that fosters and
facilitates free and open democratic debate and that promotes
positive change in the world.
We are troubled by reports that the power of Twitter was
misused by a foreign actor for the purpose of influencing the
U.S. presidential election and undermining public faith in the
democratic process. The abuse of our platform to attempt state-
sponsored manipulation of elections is a new challenge for us
and one we are determined to meet. Today we intend to show the
Committee how serious we are about addressing this new threat
by addressing the work we are doing to understand what happened
and to ensure that it does not happen again.
At the time of the 2016 election, we observed and acted on
instances of automated and malicious activity. As we learned
more about the scope of the broader problem, we resolved to
strengthen our systems, going forward. Elections continue all
the time, so our first priority was to do all we could to block
and remove malicious activity from interfering with our users'
experience. We created dedicated teams within Twitter to
enhance the quality of information our users see and to block
malicious activity whenever and wherever we find it.
Those teams continue to work every day to ensure Twitter
remains a safe, open, transparent and positive platform. We
have also launched a retrospective review to find Russian
efforts to influence the 2016 election through automation,
coordinated activity, and advertising.
While that review is still underway, we have made the
decision to review and share what we know today, in the
interest of transparency and out of appreciation for the
urgency of this matter. We do so recognizing that our findings
may be supplemented as we continue to work with the Committee
staff and other companies, discover more facts, and gain a
greater understanding of these events.
My written testimony details the methodology and current
findings of our retrospective review in detail. We studied
tweets from the period September 1st to November 15th, 2016.
During that time, we did find automated and coordinated
activity of interest. We determined that the number of accounts
we could link to Russia and that were tweeting election-related
content was comparatively small: about one one-hundredth of a
percent of total Twitter accounts at the time we studied.
One-third of one percent of the election-related tweets
people saw came from Russian-linked automated accounts. We did,
however, observe instances where Russian-linked activity was
more pronounced, and have uncovered more accounts linked to the
Russian-based Internet Research Agency as a result of our
review.
We have also determined that advertising by Russia Today in
seven small accounts was related to the election and violated
either the policies in effect at the time or that have since
been implemented. We have banned all of those users as
advertisers, and we will donate that revenue to academic
research into the use of Twitter during elections and for civic
engagement.
We are making meaningful improvements based on our
findings. Last week, we announced industry-leading changes to
our advertising policies that will help protect our platform
from unwanted content.
We are also enhancing our safety systems, sharpening our
tools for stopping malicious activity, and increasing
transparency to promote public understanding of all of these
areas. Our work on these challenges will continue for as long
as malicious actors seek to abuse our system and will need to
evolve to stay ahead of new tactics.
We have heard concerns about Russia actors' use of Twitter
to disrupt the 2016 election and about our commitment to
addressing that issue. Twitter believes that any activity of
that kind, regardless of magnitude, is unacceptable, and we
agree that we must do better to prevent it. We hope that our
appearance today and the description of the work we have
undertaken demonstrates our commitment to working with you, our
industry partners, and other stakeholders to ensure that the
experience of 2016 never happens again.
Cooperation to combat this challenge is essential. We
cannot defeat this evolving, shared threat alone. As with most
technology-based threats, the best approach is to combine
information and ideas to increase our collective knowledge.
Working with the broader community, we will continue to test,
to learn, to share, and to improve so that our product remains
effective and safe.
I look forward to answering your questions.
[The prepared statement of Mr. Edgett follows:]
[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]
Chairman Burr. Thank you, Mr. Edgett.
Mr. Walker, the floor is yours.
STATEMENT OF KENT WALKER, SENIOR VICE PRESIDENT AND GENERAL
COUNSEL, GOOGLE
Mr. Walker. Thank you very much, Chairman Burr, Vice
Chairman Warner, Members of the Committee, for the opportunity
to speak with you today. My name is Kent Walker. I'm Senior
Vice President and General Counsel at Google. I oversee our
legal, our policy, our trust and safety, and our Google.org
teams. I've worked at the intersection of technology, security,
and the law for over 25 years, starting my career as an
assistant U.S. attorney for the U.S. Department of Justice
focused on technology crimes.
Let me start my conversation with you today by joining your
earlier comments, acknowledging the victims and families of the
awful attack in New York yesterday. As a New York employer, we
know how strong and tough New Yorkers are and we look forward
to do anything we can to help.
Turning to the issues before the Committee today, Google
believes that we have a responsibility to prevent the misuse of
our platforms, and we take that very seriously. Google was
founded with the mission of organizing the world's information
and making it universally accessible and useful. The abuse of
the tools and platforms we build is antithetical to that
mission.
Google is deeply concerned about attempts to undermine the
democratic elections. We are committed to working with
Congress, law enforcement, others in our industry, and the NGO
community to strengthen protections around elections, to ensure
the security of users, and to help combat disinformation. We
recognize the importance of this Committee's mandate, and we
appreciate the opportunity to share information and talk about
solutions.
Of course, disinformation and propaganda campaigns aren't
new and have involved many types of media and publications over
the years. And for many years, we've seen attempts to interfere
with our online platforms. We take these threats very
seriously. We've built industry-leading security systems, and
we've put those tools into our consumer products as well.
Back in 2007, we launched the first version of our Safe
Browsing tool, which helps protect users from phishing,
malware, and other attacks. Today, Safe Browsing is used on
more than 3 billion devices worldwide.
If we suspect that users are subject to government-
sponsored attacks, we warn them about that. And last month, we
launched our Advanced Protection Program, which helps protect
those at greatest risk of attack, like journalists, business
leaders, and politicians.
We face motivated and resourceful attackers, and we are
continually evolving our tools to stay ahead of ever-changing
threats. Our tools don't just protect our physical and network
security, but they also detect and prevent attempts to
manipulate our systems. On Google News, for example, we use
fact-check labels to help users spot fake news. For Google
search, we have updated our quality guidelines and evaluations
to help surface more authoritative content from the web. We've
updated our advertising guidelines as well to prohibit ads on
sites that misrepresent themselves.
And on YouTube, we employ a sophisticated spam and security
breach detection system, designed to detect anomalous behavior
and catch people trying to inflate view counts of videos or
numbers of subscribers. And as threats evolve, we will
continually adapt in order to understand and prevent new
attempts to misuse our platforms.
With respect to the Committee's work on the 2016 election,
we've looked across our products to understand whether
government-backed entities were using our products to
disseminate information in order to interfere with the U.S.
election. While we did find some deceptive activities on our
platform associated with suspected government-backed accounts,
that activity appears to have been relatively limited. Of
course, any activity like this is more than we would like to
see. We've provided the relevant information to the Committee,
have issued a public summary of the results of our review, and
we will continue to cooperate with the Committee's
investigation.
Going forward, we will continue to expand our use of
cutting-edge technology to protect our users and will continue
working with governments to ensure that our platforms aren't
abused. We will also be making political advertising more
transparent, easier for users to understand, and even more
secure. In 2018, we'll release a transparency report showing
data about who is buying election ads on our platform and how
much money is being spent. We'll pair that transparency report
with a database, available for public research, of election and
ad content across our ads products.
We're also going to make it easier for users to understand
who bought the election ads they see on our networks. Going
forward, users will be able to easily find the name of any
advertiser running an election ad on Search, YouTube, or the
Google Display Network through an icon on the ad. We'll
continue enhancing our existing safeguards to ensure that we
permit only U.S. nationals to buy U.S. election ads.
We already tightly restrict which advertisers can serve ads
to audiences based on political leanings. Moving forward, we'll
go further by verifying the identity of anyone who wants to run
an election ad or use our political interest-based tools, and
confirming that that person is permitted to run that ad.
We certainly can't do this alone. We'll continue to work
with other companies to better protect the collective digital
ecosystem. And even as we take our own steps, we remain open to
working on legislation that promotes electoral transparency.
Moreover, our commitment to addressing these issues extends
beyond our services. We've offered in-person briefings and
introduced a suite of digital tools designed to help election
websites and political campaigns protect themselves from
phishing, unauthorized account access, and other digital
attacks. We're also increasing our long-standing support for
the bipartisan Defending Digital Democracy Project.
Let me conclude by recognizing the importance of the work
of this Committee. Our users, advertisers, and creators must be
able to trust in their security and safety. We share the goal
of identifying bad actors who attempted to interfere with our
systems and abuse the electoral process.
We look forward to continued cooperation both with the
Members of this Committee and with our fellow companies to
provide access to tools that help citizens express themselves,
while avoiding abuses that undercut the integrity of elections.
Thank you again for the opportunity to tell you about our
ongoing efforts. We look forward to our continuing work with
Congress on these important issues, and we are happy to answer
any questions you might have.
[The prepared statement of Mr. Walker follows:]
[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]
Chairman Burr. Mr. Walker, thank you for your testimony.
The Chair would recognize himself and share with Members
that I'm going to talk about one specific ad that--it's not
going to count to my seven minutes, and the Vice Chairman is
going to do the same at the beginning of his, to sort of set
the stage for much of what we'll talk about today.
As an example, I'd like to highlight one specific case with
real-world implications involving two different Facebook
groups, both of which are associated with the Russian Internet
Research Agency.
You'll see the first board that is up.
[The material referred to follows:]
[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]
The first group's called ``The Heart of Texas,'' with over
250,000 followers. This account promoted pro-Texas causes and
included posts many would characterize as anti-immigration or
anti-Muslim. The tagline for this group, as referenced in the
top left-hand corner of the first chart, is ``Texas: homeland
of guns, barbecue, and your heart,'' with the words ``time to
secede'' emboldened on the Texas flag.
Turning to the second group, which is in the bottom right-
hand, it's called ``The United Muslims of America,'' with over
328,000 followers. This account claimed to be pro-Islamic
themes. The tagline for this group, as referenced in the
bottom-right corner of the first chart, is ``I'm a Muslim and
I'm proud.''
So, if I could have the second board up. The Heart of Texas
group created a public event on Facebook, to occur at noon, May
21st, 2016, at the Islamic Center in Houston, Texas, to stop,
quote, ``to stop the Islamization of Texas,'' unquote. The same
group then placed an advertisement on Facebook to promote their
event, with over 12,000 people viewed.
The United Muslims of America subsequently created an event
on Facebook to occur at noon, May 21st, 2016, at the Islamic
Center in Houston, Texas, to, I quote, ``save Islamic
knowledge''--same time, same place as the Heart of Texas event.
The group then placed an advertisement, targeting people in
Houston, Texas, area to promote their event to support the
Islamic Center. More than 2,700 people viewed this ad.
If I could have the third board.
On May 21st, 2016, local news captured the events as they
unfolded, reporting on the protest stage by the Heart of Texas
group and resulting counter-protest. The pictures you see on
the third board are from the streets in front of the Islamic
Center in Houston, Texas.
What neither side could have known is that Russia trolls
were encouraging both sides to battle in the streets and create
division between real Americans. Ironically, one person who
attended stated, ``The Heart of Texas promoted this event, but
we didn't see one of them.'' We now know why. It's hard to
attend an event in Houston, Texas, when you're trolling from a
site in St. Petersburg, Russia. Establishing these two
competing groups, paying for the ads, and causing this
disruptive event in Houston cost Russia about $200.
Mr. Stretch, you commented yesterday that your company's
goal is bringing people together. In this case, people were
bought together to foment conflict, and Facebook enabled that
event to happen. I would say that Facebook has failed their
goal. From a computer in St. Petersburg, Russia, these
operators can create and promote events anywhere in the United
States and attempt to tear apart our society. I'm certain that
our adversaries are learning from the Russian activities and
even watching us today. Simply put, you must do better to
protect the American people and, frankly, all of your users
from this kind of manipulation.
My time can start now. I have one simple question, yes or
no from each of you. I'll start with Mr. Stretch and work my
way to your left.
The Federal Election Campaign Act prohibits any foreign
national from spending funds in connection with any Federal,
State, or local elections in the United States. Doesn't this
law prohibit your publication of this content?
Mr. Stretch.
Mr. Stretch. Prohibit publication of the content we've
seen?
Chairman Burr. Does FEC law apply to Facebook?
Mr. Stretch. Certainly, FEC law, yes, applies to----
Chairman Burr. Prohibits foreign dollars influencing an
election?
Mr. Stretch. It prohibits foreign actors from using really
any medium, including Facebook, to influence a foreign--a U.S.
election.
Chairman Burr. So FEC law applies to Facebook?
Mr. Stretch. Yes, it does.
Chairman Burr. Mr. Edgett.
Mr. Edgett. It applies to Twitter as well.
Chairman Burr. It applies Twitter.
Mr. Walker.
Mr. Walker. Yes, sir.
Chairman Burr. Great.
The prevalence of social media use among military members,
who spend so much time outside the country, deployed away from
friends, away from family, seems a likely target for foreign
intelligence agencies who want to collect details on U.S. force
movements, deployments, and other sensitive insight. Do you
monitor your platforms for indications that your users in the
U.S. military are targeted in any way?
Mr. Stretch.
Mr. Stretch. Senator, yes, and I would say that that sort
of--that sort of security work really falls into the
traditional cybersecurity work that we've long been focused on.
We've had a threat intelligence team for years now that has
been focused on tracking foreign actors, and it's exactly that
sort of threat that we believe has historically been an area of
focus for our adversaries, and likewise an area of focus for us
on the defensive side.
Chairman Burr. Mr. Edgett.
Mr. Edgett. Similar to Mr. Stretch, we've been focused on
that type of threat for years. We're also focused on education
on the other side and helping law enforcement and military
personnel understand how to use Twitter and both its benefits
and its risks.
Chairman Burr. Mr. Walker.
Mr. Walker. We've been looking at cyber espionage for some
years, and so this is all in focus. Because we're not a social
network, we may not have as much visibility as to whether
individual users of our service are veterans or not, but that
would certainly be an area of concern.
Chairman Burr. These questions are for Facebook, Mr.
Stretch. In a blog published September 6th, 2017, Alex Stamos,
Facebook's Chief Security Officer, wrote that the company had
discovered about 3,000 political ads that were paid for through
470 fake accounts and pages that likely operated out of Russia.
Facebook shut down these accounts on the grounds that they were
inauthentic. Had these accounts not violated Facebook's
prohibition against fake accounts, would they have been shut
down?
Mr. Stretch. Senator, many of them would have, because many
of them violated other policies related to the type of content
that's permitted on the platform. The authenticity issue is the
key.
Referring to the content you surfaced earlier, it pains us
as a company, it pains me personally, to see that we were--that
our platform was abused in this way. People in this country
care deeply in--about issues of public concern, and it's one of
the strengths of our country that people are so willing to
speak freely about them. The fact that foreign actors were able
to use our platform to exploit that openness is a deeply
painful lesson for us, and one we're focused on learning from
going forward.
Chairman Burr. Does it trouble you that it took this
Committee to get you to look at the authentic nature of the
users and the content?
Mr. Stretch. Senator, we are certainly troubled--I'd say
more than troubled--by the evidence of abuse of our platform
during 2016, and we're certainly grateful for the Committee's
investigation and the attention you're bringing to this issue.
We think it's very important.
We do believe that it's a larger issue than any one
company, and we believe that, going forward, there are
opportunities, not just for us to do better, but for us to work
together to make sure we're all addressing this threat
appropriately.
Chairman Burr. What characteristics would indicate that an
account or a page is likely operated out of Russia?
Mr. Stretch. There are a number of characteristics that can
signal potential location. The most obvious one that is
typically the most reliable is location information that's
transmitted by the user's browser when they access Facebook.
It's also the most easily manipulable.
There are many other signals that similarly will suggest
location, but, because of the way the internet is architected,
can also be faked. Our job is to look not just for the signals
that are in plain sight, but understand how they can be
manipulated, and look for patterns of activity that reveal
efforts to abuse our platform that are shrouded, both
geographically and in other ways.
Chairman Burr. Mr. Edgett, your vice president at Twitter
stated that Twitter's expanding its team and resources and
building new tools and processes to combat automated Twitter
accounts, or bots. What is Twitter's process for identifying a
bot?
Mr. Edgett. We have a lot of data behind sort of the things
you see on Twitter that looks at the activity of an account--
and remember, there are hundreds of millions of accounts--the
activity of an account as it relates to other accounts. So as
you or I, Senator, tweet, our activity looks pretty normal. As
an automated account tweets thousands of times an hour, or logs
in thousands of times a day, that looks pretty suspicious.
So our technology is looking for that anomaly that
differentiates sort of normal accounts from automated accounts.
But spammers and bad actors are getting better at making
themselves look more real.
Chairman Burr. So what percentage of accounts on Twitter
are actually bots and not real people?
Mr. Edgett. So we do a monthly audit of this and
investigation, and determined that for years less than 5
percent of our accounts are false accounts or spam.
Chairman Burr. What happens to accounts on Twitter that are
suspended by Twitter? Is there an indefinite status?
Mr. Edgett. Once we suspend an account, they're--especially
an automated account--they're typically permanently banned from
the platform. And we also do work to link those accounts with
new accounts that may pop up. So the more we investigate and
look into this and build the web of information around the
signals we're seeing from these accounts, the better we get at
linking those accounts and stopping them before they get on the
platform.
Chairman Burr. My time has expired, but I'm going to ask
you to submit in writing for the record Twitter's assessment of
why independent assessments of the number of bots on Twitter
constantly, consistently, are higher than the 5 percent that
you've stated today, if you would provide that for the record.
Mr. Edgett. Happy to provide that for the record and
address it today.
Chairman Burr. Thank you.
Vice Chairman.
Vice Chairman Warner. Thank you, Mr. Chairman.
I also want to demonstrate, but I'd also--as we're getting
ready--we have had testimony before this Committee from a
representative of NATO that fake and bot accounts on Twitter
are more in the 12 percent to 15 percent account. A vast number
of research studies--you know, 320 million active Twitter
accounts--even if you assume 10 percent, you're still talking
30-plus million potential accounts that could be misused and
abused.
If we could put up the chart here, this is another example
of how people are kind of lured in.
[The material referred to follows:]
[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]
Vice Chairman Warner. The first ad is an ad that is pretty
benign. It's obviously targeted towards Christians. It's a--
it's an ``Army of Jesus'' Facebook ad, 217,000 followers. You
like that page, and here's what happens: you would get a series
of, for the most part, relatively benign Bible quotes or other
items. This ad appeared in October of 2016.
Late October, early November, suddenly, this benign site,
in addition to your Bible quotes, suddenly we're getting these
other posts, not paid ads, but posts from this organization.
This message, obviously the bottom one would've gone to the
217,000 followers. We have no idea how many times it was liked
or shared with other individuals.
Again, we've got two different examples of the type of
tools, how people are lured in, and then once they're lured
into what they think is a pro-Texas or pro-Muslim or here pro-
Jesus account--and then they are manipulated by foreign actors
and foreign agents.
Go ahead and start my time.
First of all, I hear all your words, but I have more than a
little bit of frustration that many of us on this Committee
have been raising this issue since the beginning of this year,
and our claims were frankly blown off by the leaderships of
your companies, dismissed, said, there's no possibility,
nothing like this happening, nothing to see here.
It bothers me, if you're really committed to trying to work
with us to resolve this, that it took until this Committee
continually went at you, and it was July and early August when
you made your first presentations. And, candidly, your first
presentations were less than sufficient and showed in my mind a
lack of resources, a lack of commitment, and a lack of genuine
effort.
Candidly, your companies know more about Americans in many
ways than the United States government does. And the idea that
you had no idea any of this was happening strains your
credibility.
So my first question is this, and I want a yes or no
answer, not a filibuster. Will you commit to continue to work
with this Committee to provide additional information and
additional documents as needed as we continue to explore this
challenge and threat on a going-forward basis?
We go right down the line. Mr. Stretch.
Mr. Stretch. Yes.
Vice Chairman Warner. Mr. Edgett.
Mr. Edgett. Absolutely.
Vice Chairman Warner. Mr. Walker.
Mr. Walker. Absolutely.
Vice Chairman Warner. Next, one of the things that I
continue--again, and I will commend you here--that from the
first our friends at Facebook, you identified 470 accounts,
3,000 ads. And most of the work, at least it appears to me,
from at least Twitter and Facebook, has all been derivative of
that initial data dump.
And again, this is a yes or no question: Do you believe
that any of your companies have identified the full scope of
Russian active measures on your platform? Yes or no?
Mr. Stretch. Senator, our investigation continues. So I
would have to say no, certainly not with certainty.
Vice Chairman Warner. Mr. Edgett.
Mr. Edgett. No, and we're still working on this.
Vice Chairman Warner. Mr. Walker.
Mr. Walker. I believe we haven't done a comprehensive
investigation, but, as Mr. Stretch says, these are ongoing
issues, and we continue to investigate.
Vice Chairman Warner. Let me start again with Facebook
here. You've identified 470 accounts from one troll farm in St.
Petersburg. There have been plenty of press reports of other
troll farms in Russia. There have been reports of other
activities that were Russian-controlled in Central Europe and
Eastern Europe. In meetings with your leadership, as you became
more aware of this problem, you aggressively promoted the fact,
for example, that you took down 30,000 accounts around the
French elections. Now, you say not all of those were Russian-
related.
Have you gone back and cross-checked those Russian-related
accounts that you took down in France to see if any of those
accounts were active in the American election?
Mr. Stretch. Senator, the 30,000 accounts that we took down
in----
Vice Chairman Warner. The accounts that were related to
Russian accounts that you took down. Your leadership in fact
bragged about how proactive you were in the French election
process. Did you check those accounts to see if any of them
were active in the American elections?
Mr. Stretch. Senator, the system that ran to take down
those accounts, which were fake accounts of, really, all type
and for any purpose, has--is now active worldwide----
Vice Chairman Warner. Have you----
Mr. Stretch [continuing]. And has been operated----
Vice Chairman Warner. Just please answer my question. Have
you reviewed the accounts you took down in France that were
Russian-related to see if they had played any role in the
American election?
Mr. Stretch. Senator, I apologize. I'm trying to answer the
question.
Vice Chairman Warner. Well, the answer is yes or no. I
don't want a long explanation. I want to know if you have done
this. I've been signaling this to you for some time. We wanted
to make sure that you would review those accounts. We wanted to
make sure--the 470 accounts that paid for the 3,000 ads, you
said these were all accounts, except for one, that were paid
for in rubles. Did you even run those accounts to see if any of
those accounts were paid for with dollars or euros or other
currencies?
Mr. Stretch. Senator, those--let me try to state it----
Vice Chairman Warner. Yes or no?
Mr. Stretch. So we have, and we continue----
Vice Chairman Warner. Mr. Stretch, yes or no?
Mr. Stretch. Yes, we are looking and have looked at every
possible indication of Russian activity in the 2016 election,
and the investigation continues.
Vice Chairman Warner. Sir----
Mr. Stretch. That includes any evidence we've identified
from those 30,000 accounts, as well as a number of----
Vice Chairman Warner. All those accounts have been run,
that database has been run, to see if any of those accounts
were active in the United States?
Mr. Stretch. I will have to come back to you on that,
Senator.
Vice Chairman Warner. Sir, we've had this hearing scheduled
for months. I find your answer very, very disappointing.
On the question of--we just discovered, and I appreciate
this, you had 80,000 views in terms of Russian--views on
Facebook. We now discovered in the last 48 hours 120,000
Russian-based posts on Instagram. Have you done any of the
similar analysis on those 120,000 posts? We know the 80,000
ended up reaching 126 million Americans. Have you done that
same analysis on the 120,000 posts on Instagram?
Mr. Stretch. Yes, Senator, we have.
Vice Chairman Warner. And how many Americans did those
touch?
Mr. Stretch. The data on Instagram is not as complete, but
the data that we do have indicates that, beginning in October
of 2016, those Instagram posts reached an additional 16 million
people, in addition to the 126 million people that we
identify----
Vice Chairman Warner. So now we're seeing the Russian
activities roughly at 150 million-plus Americans, without
knowing how many times they were re-shared.
Mr. Stretch. If I can add that the time period prior to
October 16th, where our data is less reliable, would yield an
incremental 4 million. So all told, that gets you to
approximately a little less than 150 million. That's correct,
Senator.
Vice Chairman Warner. Mr. Edgett, on the Twitter account,
you--there was one activity--and this was not something that
happened during 2016. Again, I agree with the Chairman. We're
not here to relitigate 2016. But there was a fake Tennessee
Republican account, TEN-GOP. The irony was this account had
154,000 followers; the real Tennessee GOP party had 13,000--
13,400 followers, I believe, based on your numbers.
I find very interesting, there have been some people who
have said, ``Well, people should be able spot these fake
accounts.'' Well, if they're able to spot these fake accounts,
you had the President's communications director retweeting this
account, Kellyanne Conway. You had the President's son, Donald
Trump, Jr., retweeting this account.
My question is, why did it take so long to take this down
when the Tennessee Republican party was asking you repeatedly?
Mr. Edgett. Yeah, that--and that was an absolute miss, and
we've gotten better since. We've refined our policies around
impersonation and parody accounts----
Vice Chairman Warner. Let me just close with this. My
time's about up.
Mr. Edgett. Sure.
Vice Chairman Warner. We've looked on this subject, on
political information and disinformation. But the same way that
these bots and trolls and click farms and fake pages, groups,
algorithm gaming can be used in politics, these same tools can
and have been used, I believe, to assist financial frauds
around stock schemes. I think there is a lot of this activity
in broad-based digital advertising. I think we've seen some of
this in schemes to sell counterfeit prescription drugs, as well
as the ability to encourage folks to download malware. I
believe this is a real challenge, and to get this right we're
going to need your ongoing cooperation.
Thank you, Mr. Chairman.
Chairman Burr. Senator Risch.
Senator Risch. Thank you.
Gentlemen, thank you for coming today. By now it's probably
pretty obvious to everyone that this Committee has spent lots
and lots and lots of time on this, both as it relates to the
election and on these kinds of things not related to the
political process here in the country. But we have spent a lot
of time, and I think have been able to reach some conclusions
on this.
No-one's exempt. I come from a State that's a lot smaller,
Idaho. They tried to do exactly in Idaho exactly what was done
in Texas, where they tried to promote a meeting where they had
two conflicting sides, and no one showed up. So there was no
success.
In Idaho, just like Texas, it had absolutely nothing to do
with the 2016 presidential election. It was simply a cultural
type of acrimony that they were attempting to promote.
The Chairman talked about the news reports that allege that
the Russians used social media to promote a particular
candidate, and may even, some of those, suggest that it changed
the result of the election. But this whole thing goes a lot
deeper than that. One of the things we've discovered--and I
think you probably are aware of this--that you can't look at
those ads and say, ``Okay, they were all promoting one
particular candidate.'' There were ads going both ways, for and
against both candidates, by the Russians. I'm going to get back
to that in just a second.
But the other thing that I think that we've come to a
conclusion on, and very early, is that the U.S. isn't the only
country suffering from this. The Europeans, France, Austria,
Germany, just among others, have suffered from the exact same
thing, and that is Russian attempted interference with their
domestic affairs.
I put a section in the sanctions bill, the Iran, Russia,
North Korea sanctions bill, that requires the Executive Branch
to do a study on this, on the effect in Europe, because they
were much more overt in Europe than they were here--most of the
work they did here was covert--and probably because in European
countries there is actually a fair amount of Russian sympathy
where they can mobilize these people, not so much here in the
United States. Obviously some, but not nearly, what is there.
So I want to come at this from a different perspective. The
2016 elections got a lot of the politicians riled up because it
went after the political process. But my conclusion is, and I
think most people here would agree with me, that--and indeed,
Senator Warner referred to this--that this is a lot deeper than
just the elections.
There are a lot of things that the Russians are trying to
do, and not just inject themselves into the electoral process.
It seems to me that, after you step back and look at this and
say, ``What's going on here? What is the motivation? What are
they doing?'' I always look at something from an objective
standpoint--``What is their objective? What are they trying to
accomplish?'' And you walk away from it just shaking your head,
because we Americans don't think the same way they think about
promoting our country.
So the conclusion I've reached is that the Russians are
doing what they've done all along, long before your technology
even existed, and that is trying to sow discord, simply trying
to sow discord.
My question to each of you is: have you tried to analyze
what the Russians were trying to accomplish here, not only in
the 2016 elections, but in these other kinds of ads, with the
discord? What are your personal views on that, whether
they've--what they're trying to accomplish?
Mr. Stretch.
Mr. Stretch. Senator, it's very difficult for us to ascribe
motive. It's I think why this Committee's work is so important.
We've tried to provide you as much information as we can, and
we hope that, with your visibility into other sources of
information, you will be able to help the American people have
a better assessment of what the motive is. We think that'll
help all of us do better to prevent this sort of activity in
the future.
Senator Risch. Would you agree with me that the motive
isn't obvious here, given the difference in the way they handle
these things?
Mr. Stretch. Yes, I would agree with that.
Senator Risch. Mr. Edgett.
Mr. Edgett. I would agree with that, as well. I mean, based
on what we've seen, the advertisements from Russia Today, the
types of content that was being put out by the IRA, also the
automated account content, looks as if it's merely focused on
divisiveness. But we're still investigating this issue, and
look forward to working with this Committee to help put the
whole picture together.
Senator Risch. Mr. Walker.
Mr. Walker. The large majority of the material we saw was
in the socially divisive side, rather than direct electoral
advocacy, yes.
Senator Risch. And that has really been the focus of the
media, that, oh, this was all about the 2016 election. You
agree with me that this is much broader than that and is, as
you say, divisive--aimed at divisiveness, or aimed at discord?
Would you all agree with that?
Mr. Edgett. Yes, and that's a problem we're trying to
tackle every day.
Senator Risch. Mr. Stretch, do you agree with that?
Mr. Stretch. Yes, I would agree and note that the time
period in question and the activity we saw even continued after
the election.
Senator Risch. Mr. Walker.
Mr. Walker. That seems reasonable, hard for us to know and,
again, ultimately for the Committee to decide.
Senator Risch. I appreciate that. And as I said, my view of
this is this a whole lot broader than simply the 2016 election.
Mr. Walker, I have a specific question for you. I think I
heard you say that you're enacting a policy where only a U.S.
national can buy an election ad. Is that correct?
Mr. Walker. That's correct.
Senator Risch. Okay. What about other countries? Obviously,
you operate in places other than the United States. Can a U.S.
national buy an ad, for instance, for a French or German or
Austrian campaign?
Mr. Walker. I haven't studied the laws of individual
countries, but we are not confining our work to the U.S. We are
looking at other elections around the world to make sure that
we do whatever we can to minimize electoral inference.
Senator Risch. So what you're going to do is try to confine
people to their own elections in their own countries? Is that
pretty much your objective?
Mr. Walker. Certainly that's the case for the United
States, and any other country around the world where that's the
law that's true, yes.
Senator Risch. I think that's going to be a big challenge
for you, but good luck, and I wish you well in that endeavor.
Thank you, Mr. Chairman.
Chairman Burr. Senator Feinstein.
Senator Feinstein. Thanks, Mr. Chairman.
I sat in the Judiciary hearing yesterday. It was a
subcommittee hearing, but was able to ask some questions. And I
want to just make a personal comment, because I've been very
proud, and I know Senator Harris is as well, to represent this
tech community from California. But I must say, I don't think
you get it--I think the fact that you're general counsels, you
defend your company--that what we're talking about is a
cataclysmic change. What we're talking about is the beginning
of cyber warfare. What we're talking about is a major foreign
power with the sophistication and ability to involve themselves
in a presidential election and sow conflict and discontent all
over this country.
We are not going to go away, gentlemen, and this is a very
big deal. I went home last night with profound disappointment.
I asked specific questions. I got vague answers, and that just
won't do.
You have a huge problem on your hands, and the United
States is going to be the first of the countries to bring it
your attention, and others are going to follow, I'm sure,
because you bear this responsibility. You've created these
platforms and now they are being misused, and you have to be
the ones to do something about it, or we will.
And this Committee is Intelligence. It's different from
yesterday, so they're privy to different facts, and they're
very potent facts.
Let me go back to a couple of questions that I asked
yesterday. Mr. Edgett, yesterday, you testified that Twitter
only began to remove voter suppression posts that told people
they could vote by texting or tweeting after you found out
about them from other Twitter users.
These were illegal tweets. Waiting for users to alert
Twitter isn't sufficient. I'll give you another chance. What is
Twitter doing to proactively identify illegal voter suppression
tweets?
Mr. Edgett. Thank you for letting me address that. We're
constantly improving, not only on our technology around
automated accounts that are trying to amplify these types of
messages----
Senator Feinstein. That's not enough.
Mr. Edgett [continuing]. But also on putting people and
technology on the content and the behavior and trying to make
our workflows, our reporting flows, more efficient and using
artificial intelligence to prioritize things like the illegal
voter suppression ads and other things that we see on the
platform, and taking those down faster.
We are getting better, but this is a problem that we are
focused on getting better at every day.
Senator Feinstein. Well, you have to find a way to prevent
them from going up.
Mr. Edgett. That's right. And that's why we tend to----
Senator Feinstein. That's the problem.
Mr. Edgett. Right. That's why we tend to focus on behavior
behind the accounts, to know before the content goes up. We've
seen--we've seen great strides in other areas not related to
that, and so we're trying to take that same solution to this
problem set.
Senator Feinstein. Mr. Walker, I asked your colleague
yesterday why Google didn't immediately revoke Russia Today's
preferred status after the intelligence community determined
and publicly stated that RT was a part of the Russian
government's efforts to interfere in our election.
Mr. Salgado told me that RT only lost its preferred status
because of a, quote, ``drop in viewership,'' end quote, not
because it was part of the Kremlin's propaganda machine. This
response was deeply troubling, and frankly, did not answer my
question. So here it is again. Why didn't Google take any
action regarding RT after the intelligence community assessment
came out in January of 2017?
Mr. Walker. Let me start by----
Senator Feinstein. I'm sorry, your mic isn't on.
Mr. Walker. Senator, let me start by responding to your
initial comments to assure you we take this and have taken this
issue very seriously. The question of cyber espionage is one
that we have been working on for some years, publicly and
privately, working with other companies and working on our own
to identify some of these threats. This is one manifestation of
that, but not the only one.
With regard to RT, we recognize the concerns that have been
expressed about RT and concerns about its slanted coverage.
This is of course a question that goes beyond the internet. RT
is covered--its channel is on major cable television stations,
on satellite television stations. Its advertising appears in
newspapers, magazines, airports. It's run in hotels in pretty
much every city in the United States.
We have carefully reviewed the content of RT to see that it
complies with the policies that we have against hate speech,
incitement to violence, et cetera. So far, we have not found
violations, but we continue to look.
Beyond that, we think that the key to this area is
transparency--that Americans should have access to information
from a wide variety of perspectives, but they should know what
they're getting. And so we already on Google provide
information about the government-funded nature of RT. We're
looking at ways to expand that to YouTube and potentially other
platforms.
Senator Feinstein. Well, as you might guess, I'm really not
satisfied with that. That's been the trend of the testimony all
along. I think we're in a different day now. We're at the
beginning of what could be cyber war. And you all, as a policy
matter, have to really take a look at that and what role you
play.
I think my time is almost up. Let me try one more. A
British report recently concluded that social media platforms
such as Facebook, Twitter, and YouTube failed to remove
extremist material posted by banned jihadist and neo-Nazi
groups, even when that material was reported. The source for
this is the British Parliament Home Affairs Select Committee.
Last night, we saw a horrific attack on innocent people in
New York by an individual who may have been radicalized online.
We know one person, who is Anwar al-Awlaki, with 75,000 hits,
the major radicalizer in the United States on the internet.
I'm working on legislation to require tech companies to
report known terrorist activity on their platforms to law
enforcement and to provide law enforcement with civil
injunction authority.
So thank you, Mr. Chairman.
Chairman Burr. Thank you, Senator Feinstein.
Senator Rubio.
Senator Rubio. Thank you.
Thank you all for being here.
Mr. Stretch, I want to ask you--and it relates to all this
in the following way, but let me work there. Guo Wengui is a
whistleblower and a critic of the Chinese government, and his
Facebook account was blocked, and Facebook has informed us and
has said publicly that he violated terms of service. I think he
published personal identifying information about individuals,
and that violated the terms of service, so--and I understand
that argument.
My question--so what I want to be clear is, was there any
pressure from the Chinese government to block his account?
Mr. Stretch. No, Senator. We reviewed a report on that
account and analyzed it through regular channels using our
regular procedures. The blocking was not of the account in its
entirety, but I believe was of specific posts that violated our
policy.
Senator Rubio. But you can testify today that you did not
come under pressure from the Chinese government or any of its
representatives, or people working for them, to block his
account or to the block whatever it is you blocked?
Mr. Stretch. I want to make sure I'm being precise and
clear. We did receive a report from representatives of the
Chinese government about the account. We analyzed that report
as we would any other and took action solely based on our
policies.
Senator Rubio. Facebook is not allowed to operate in China.
Is that correct?
Mr. Stretch. Yes, that's correct. Our consumer services are
blocked in China, that's correct.
Senator Rubio. Okay. There have been press reports that
Facebook may have potentially developed software to suppress
posts from appearing in people's news feeds in specific
geographic areas. And the speculation is it's being done for
the purposes of getting into the Chinese market.
Is that accurate? Has Facebook developed software to
suppress posts from appearing in people's news feeds in
specific geographic areas?
Mr. Stretch. Senator, as you know, we are--we are blocked
in China, so any software we have is certainly not operative
there. We do have many instances where we have content reported
to us from foreign governments that is illegal under the laws
of those governments. So a great example of this is Holocaust
denial in Germany, for example. And our position with respect
to reports like that is, if there is content that's visible in
a country that violates local law and we're on specific notice
of that content, we deploy what we call geoblocking, or I.P.
blocking, so that the content will not be visible in that
country, but remains available on the service.
Senator Rubio. So, for example, if criticizing a government
is illegal in that country, you have the capability to block
them from criticizing the government and thereby gaining entry
into that country and being allowed to operate?
Mr. Stretch. We have the capability to ensure that our
service complies with local law, that's accurate. We take a
very nuanced approach to reports of illegal content. We believe
our mission is to enable people to share and connect, and we
believe that political expression is at the core of what we
provide. And so----
Senator Rubio. What if that political expression is illegal
in the country?
Mr. Stretch. So, in the vast majority of cases where we are
on notice of locally illegal content, it has nothing to do with
political expression. It's things like blasphemy in parts of
the world that are--that prohibit blasphemy.
Senator Rubio. We could probably do a whole hearing on that
topic. But here's why that's related to what we're talking
about today: terms of service is the reason why he was knocked
off. All of your companies have terms of service.
Is a foreign influence campaign a violation of the terms of
service of any of the three companies represented here today?
If you can prove that someone is doing it on behalf of a
foreign government, seeking to interfere in an election, does
that violate your terms of service?
[Pause.]
Any of you? Any of the three companies, in terms of being
able to operate or post things, and particularly Twitter and
Facebook?
Mr. Edgett. Generally, it would violate a number--we don't
have state-sponsored manipulation of elections as one of our
rules. But generally the other types of rules, like
inflammatory ads content, would take down most of these, these
posts. So we don't outright ban it, but----
Senator Rubio. Well, let me ask you this. I've read that
you can buy a bot army from between $45 to $100. Is buying
and--if you can prove that someone's bought up and put together
a bot army, would that be a violation of terms of service?
Mr. Edgett. Those would violate our terms of service around
the use of automated accounts, and those are the things that
we're catching every day. We're blocking 450,000 suspicious
logins a day. We're challenging 4 million accounts every week,
to make sure that they're actually real people. But we have--we
have terms of service around----
Senator Rubio. I didn't get an answer on the face--is that
a violation of terms of service, to buy for a foreign influence
campaign or to put together a bunch of fake ads and put them
together?
Mr. Stretch. That campaign violates our terms and our
policies in a number of ways. And we do not permit automated
means for accessing the site, so using the bots likewise would
be a violation.
Senator Rubio. Okay. If someone goes on and posts the
Social Security number and date of birth of an individual,
that's a violation of terms of service, correct?
Mr. Stretch. For Facebook it is, yes.
Senator Rubio. I would imagine for all the platforms.
Mr. Edgett. It is.
Senator Rubio. What about if someone goes online and posts
classified information, illegally obtained, that threatens the
lives of individuals or methods, or potentially disrupts the
ability to disrupt a plot that can endanger the lives of
people? Is posting that online or posting that in one of your
platforms a violation of terms of service? It happens
sometimes. I don't know if you're aware of it.
Mr. Edgett. We work with law enforcement all the time on
matters like that, and balance free speech rights, obviously,
with those of--obviously, an imminent threat, we would--we
would take very seriously and act on right away.
Senator Rubio. I guess my point is, personal identifying
information is illegal to post it, right? It threatens
someone's identity. It's also illegal to steal and reveal
classified information. And I'm just curious if that's also a
violation of terms of service, since in fact it could have
real-life implications on individuals who could be compromised
because of that release.
Do we have any evidence that Russian accounts uploaded U.S.
voter registration data and used it in conjunction with custom
audiences to target specific voters by name? Do any of you have
any information that registered voter data was uploaded and
used to customize advertising or messaging to individual
voters?
Mr. Edgett. We haven't seen evidence of that.
Mr. Stretch. The same is true for Facebook.
Senator Rubio. And my last question is--the scope of this
was not limited to 2016 or even the presidential race. As an
example, I think, with the help of some of the companies here,
we've identified Being Patriotic, LGBT United, United Muslims
of America, Stop A.I., Heart of Texas; all were used to attack
my campaign during the primary. What's interesting, though, is
on the 3rd of July and on the 8th of August, after the primary
but when I chose to run for reelection, one of those, LGBT
United, attacked again.
So my point being these operations, while we're talking
about the 2016 presidential race, they're not limited to 2016,
and they were not limited to the presidential race, and they
continue to this day. They are much more widespread than one
election. It is about our general political climate. Is that
correct?
Mr. Stretch. I would certainly agree with that statement,
Senator.
Senator Rubio. Okay. Thank you.
Chairman Burr. Senator Wyden.
Senator Wyden. Thank you, Mr. Chairman.
With the current fascist leadership of Russia
enthusiastically undermining our democracy, America must defend
the values that made us great and aggressively confront this
espionage and the enemies that sponsor it. The tools of the
espionage range from political ads, to issue ads, sock puppets
to fictional news stories, and from rallies, to protests, to
marches, all presented under false pretenses.
While the Supreme Court has ruled that Congress may place
some limits on strictly political advertising, the other
activities I just mentioned are beyond the reach of government
and government regulation in a free society.
To fight back against this espionage, Americans have to
rely on our marketplace of ideas and the institutions that
support it. Gentlemen, today you three represent those
institutions.
Now, you've discussed your response to these attacks, but
it is self-evident, in relation to the power your platforms now
have, in the past election you failed. And this is especially
troubling because the same Federal law that allowed your
companies to grow and thrive, the Section 230 law, gives you
absolute legal protection to take action against those who
abuse your platforms to damage our democracy.
The same algorithms that power your companies can be used
to identify the behavior indicative of these attacks, including
fake accounts and fake news stories, and identify the source of
money purchasing your ads.
Now, I'm of the view ads are a small part of a much bigger
problem. Fake users posting stories on Facebook, videos on
YouTube, links on Twitter can be used by foreign and domestic
enemies to undermine our society. You need to stop paying lip
service to shutting down bad actors using these accounts.
You've got the power and Congress has given you the legal
protection to actually act and deal with this.
So I want to start with a couple of quick yes or no
questions and just go right down the row for the three of you.
Mr. Walker, are you satisfied with your platform's response to
foreign interference in the 2016 election? Yes or no? Just yes
or no.
Mr. Walker. We--we are constantly doing better.
Senator Wyden. Is the answer no?
Mr. Walker. We could have done more, but I think we are
doing more today and have done more since the election.
Senator Wyden. I'll take that as a no.
Mr. Edgett.
Mr. Edgett. No, we need to do more.
Senator Wyden. Mr. Stretch.
Mr. Stretch. The same is true.
Senator Wyden. Okay.
Do you all have--and we'll start with you, Mr. Walker--the
technical ability and resources to better respond to future
misinformation campaigns? Yes or no?
Mr. Walker. Yes. The safe harbors and the Good Samaritan
laws are important underpinnings for all of this. And we are
doing more, we have done more to combat fake news----
Senator Wyden. Mr. Edgett, yes or no?
Mr. Walker. Yes.
Mr. Edgett. Yes.
Senator Wyden. Mr. Stretch.
Mr. Stretch. Yes, and I would add, though, that I do
believe we need information-sharing among industry, as well as
working with the government, to enable us to do this
effectively.
Senator Wyden. All right.
Gentlemen, specifically now describe the changes you're
going to pursue that respond to not just the ads, but the sock
puppets, the hoaxes, and the confidence operations? We'd like
to walk out of here knowing the changes you're going to support
going forward.
Mr. Walker.
Mr. Walker. Sure. Let me give you four on the ad side and
three on the non-ad side.
Senator Wyden. Quickly.
Mr. Walker. Absolutely. The transparency report that we
talked about for ads; an archive of content that--of all ads'
content that's available; icons that make information on the
site available to users as to who sponsors an ad; and enhanced
verification techniques.
When it comes to non-ads material, fake news, we're
improving our algorithms, our rater guidelines, and the signals
we use. We're using fake news fact-check labels to improve
users' ability to evaluate fake news, and we're looking at our
ads policies to improve and toughen rules against sites that
misrepresent their nature.
Senator Wyden. Mr. Edgett.
Mr. Edgett. Coming out of the 2016 election and early this
year, our CEO asked our entire engineering product and design
teams, which make up a large majority of the company, to tackle
the problem of safety, abuse, and misinformation on our
platform, and to drop everything else that we're doing and to
figure this out. We formed we call an information quality team.
Senator Wyden. Those are three sentences. What are the
changes?
Mr. Edgett. Yes. We formed an information quality team
focused on looking at both behavior and content and seeing how
we could stop bad actors from using automated activity to
amplify their message. We have just announced new transparency
rules around not just political ads, but all advertisements, to
educate not just American citizens, but our worldwide users.
We are also continuing to collaborate with law enforcement
and committees like this to make sure we're putting the right--
--
Senator Wyden. I know very few specifics from that answer.
Mr. Stretch.
Mr. Stretch. Senator, let me try four things.
First, today there are 10,000 people working at Facebook on
safety and security across our security product and community
operations teams. By the end of 2018, there will be more than
20,000.
Second, we announced last week a series of ad transparency
steps, drawing on the ideas in the Honest Ads Act that Senator
Warner talked about earlier, that will bring much greater
visibility to advertising generally and particularly to
political advertising.
Third, we are tightening our ad policies to limit
divisiveness and to limit violence in the use of our ad tools.
And fourth, we're standing up an organization to enable
better industry sharing of threat information and also to help
us work better with law enforcement so that we can share
information and expertise in order to address this threat going
forward.
Senator Wyden. My last question is, it's not clear that you
all or the public understand the degree of this sophisticated
and manipulative intelligence operation. The Russians created
Facebook pages, posted YouTube videos, all trying to appeal to
specific audiences.
Some of the content wasn't fake. It was intended to gather
an audience and gain trust. It told people that they were
already receptive to that, after gaining that trust, you could
execute the espionage, for example, by gathering liberals and
then discouraging them from voters.
Mr. Stretch, I'd like you to confirm that this technique
was used in the election.
Mr. Stretch. Senator, we've provided all the information we
can about the content that we've identified on the system. I
think to make the sort of assessment you're describing really
requires this Committee's work to look at all of the online and
offline activity that would be necessary to effectuate a
campaign like that.
Senator Wyden. My time has expired. We have specific cases
that that was used. I would like in writing, within a week,
what you're doing about it.
Thank you, Mr. Chairman.
Chairman Burr. Senator Collins.
Senator Collins. Thank you, Mr. Chairman.
It is very clear that Russian activities on your social
media platforms go far beyond the paid political ads that
appeared last year. The primary purpose of Russia's active
measures is to exploit and to aggravate the divisions in
American society and to undermine public confidence in our
democratic institutions. And those efforts have not stopped.
They continue to this very day. As Senator Risch has pointed
out, no area of the country is immune.
So let me give you an example, and we've passed it out to
you, by describing three unpaid posts from Facebook pages
created by the Russians that refer to the governor of Maine,
Paul LePage.
[The material referred to follows:]
[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]
Senator Collins. There are two negative posts related to
the governor on one Russian Facebook page, called Williams &
Calvin, that appeared in August of 2016. There's a video of
comments made by Maine's governor from that same month. And the
post in part says the following: ``LePage called up white
people to kill blacks. After this statement, we can clearly see
what kind of people serve in American government: white racist
supremacy--that's for sure. The only way to avoid mass killings
of black people is to fire LePage and all who have the same
racist beliefs from American government.''
There was a second post on the same website about 10 days
later. Let me just read part of that: ``It is not a secret that
America is the country of white supremacy, and people like
LePage must be replaced from their positions in the government.
America doesn't need racist politicians. Black people are tired
of white supremacy.''
Then, this year, this year, in August of 2017, Maine's
governor was the subject of a positive post on a different
Russian-backed Facebook page, called Being Patriotic. In this
case, the post defended comments that the government made at
the time about Confederate monuments.
The post ends with its own incendiary conclusion. It says:
``When even the governor is not safe from leftist haters, then
what can we say about ordinary citizens? Liberals are now
acting like terrorists. They try to intimidate everyone who
disagrees with them. Hope our police will take appropriate
measures against these cowards.''
Now, let me point out something. Our governor was not up
for reelection last year; he is term-limited. He cannot run for
reelection as governor. And yet these comments were made both
last year and just a few months ago.
And the posts are just three among 80,000 that reveal the
Russian playbook of playing both sides off against each other
and of sowing discord and division with inflammatory rhetoric.
And there were other posts that involved lower-level officials
in the State of Maine that we found as well. And the Russians
continue to push this kind of divisive rhetoric to this very
day.
So my question to you is: what are you, as American
companies, doing to effectively counter unpaid content posted
by the Russians that is clearly designed to specifically
polarize and anger the American people? And I would argue that
you have a special obligation here, given your reach in
American society and the fact that you are patriotic American
companies.
Mr. Stretch.
Mr. Stretch. Senator, we agree that we have a special
responsibility here. We value the trust that users place in our
services. And when they show up to connect with friends and
family and to discuss issues, they need to know that the
discourse they see is authentic.
What is so painful about this type of content is that it
exploits truly and passionately held views and then inflames
them to create more discord and more distrust.
To prevent this, we are investing much more heavily in
authenticity. We believe that one of the cornerstones of
Facebook is that users are known by their real names and so
that creates a level of authenticity in the discourse that
users can trust when they come to the platform. This sort of
content erodes that trust and it's contrary to everything we
stand for as a company. As Americans, it's particularly painful
because it is so exploitative of the openness of our society.
And so the investment we are making and the commitment we
are making is to ensure that our authenticity policy is more
effectively policed and monitored to prevent exactly this sort
of behavior.
Senator Collins. Mr. Edgett, what is Twitter doing?
Mr. Edgett. We're focusing on a number of things. The one
we see the greatest strides in and where we see the greatest
effect and protections from our users is on the amplification
side, in the use of automated accounts. These bad actors need
an audience for their voice, and generally they don't have a
followership. So they are trying to use activity on the
platform to automate and amplify their voices.
So we're looking behind the message and behind the content
at the behavior of doing that, and have been successful in
doubling our effectiveness of doing that, year over year, and
looking at the behavior, taking down millions of accounts every
single week because they're not actually humans, they're
actually----
Senator Collins. Well, this just happened in August of this
year. This isn't something old.
Mr. Edgett. Right. We continue to try to stay ahead of
their activities. We're also looking at things like coordinated
human activity, where real people are coming together, like the
IRA, and actually putting out divisive content like this. We
are able to link those accounts and take action on them as we
learn, not just what they're saying, but what's behind. What's
behind it only we can see on the Twitter side. We've had great
strides on the terrorism front in that regard, and we believe
we can apply the same techniques and methodologies to this,
this problem.
Senator Collins. Mr. Walker. Thank you.
Mr. Walker. We're also very concerned about this kind of
deceptive and divisive content. We remove it immediately from
our services and we have removed these.
Going forward, and actually already, we have engaged a
number of things to avoid the problem of fake news: changes to
our algorithms, improving the training that our raters get in
evaluating quality, labeling fake news where we can find it,
working with third parties, et cetera.
Senator Collins. Thank you.
Chairman Burr. Senator Heinrich.
Senator Heinrich. Thank you, Chairman.
Mr. Stretch, I want to start with you. Last month,
President Trump called Russian-purchased Facebook ads a hoax.
I've looked at those Russian-sponsored Facebook ads. I
certainly hope you've had a chance to review them. Are they in
fact a hoax?
Mr. Stretch. All the information we've provided to the
Committee did run on Facebook, so----
Senator Heinrich. It's a yes or no answer. I know you're a
lawyer; it's hard. But----
[Laughter.]
Mr. Stretch. No. The existence of--those ads were on
Facebook and were not a hoax.
Senator Heinrich. So in the interest of just clearing this
up and giving the American people some transparency into this,
so that they can see the nature of what typically gets used to
divide the American populace, why not simply release those
Russian-financed Facebook ads to the public? Redact the
pictures, but release the contents, so that people can
understand how this works?
Mr. Stretch. Senator, we believe this Committee is really
best placed to determine what information to release. We stand
ready to assist in that, in that effort. We agree that the more
people can see the type of content that ran and the divisions
that were sought to be exploited, the better.
Senator Heinrich. Well, I think we have a disagreement on
this Committee as to whether or not to release those. I would
urge all of you as platforms to consider that kind of activity
as well.
I want to move on to Russia's RBC magazine, which recently
revealed that St. Petersburg's troll factory employed hundreds
of trolls, including 90 at the, quote-unquote, ``U.S. desk''
alone, and spent about $2.3 million in 2016 to meddle in U.S.
politics, actually contacted U.S. activists directly and
offered them thousands of dollars to organize protests.
Your platform--your platforms are all global. They're not
just U.S. platforms. And there is substantial open-source
reporting right now suggesting that similar divisive activity
may be occurring, for example, in the Catalonian region of
Spain right now.
What are each of you doing right now to make sure that your
platforms aren't being used in similarly divisive ways across
the globe, to sow discord in Western democracies? And in
particular with the Catalonian example, are you familiar with
what you're doing there?
Mr. Stretch. Senator, we are focused on preventing this
form of abuse globally. So when we say we have an obligation to
protect the platform from being used for abuse, that's a global
obligation. So we are focused on elections as they appear on
the calendar, including the Catalonian election that occurred
recently, as well as the other elections that are on the
calendar going forward.
We're focused on ensuring that all actors on the platform
comply with local law, as Mr. Walker suggested earlier, and we
are focused on making sure that any foreign threat actors that
are seeking to undermine democracy anywhere are removed from
the platform.
Senator Heinrich. Have each of you taken, had to take
corrective action against actors in that debate who are not who
they purported to be?
Mr. Stretch.
Mr. Stretch. Senator, the key I'd say progress we've made
is----
Senator Heinrich. That's a yes or no, once again.
Mr. Edgett.
Mr. Edgett. I believe so, but I'll need to follow up with
your staff.
Senator Heinrich. Thank you.
Mr. Walker.
Mr. Walker. We're constantly removing fraudulent and
deceptive accounts from our services. I'm not familiar with the
specifics there.
Senator Heinrich. You can get back to us.
Mr. Edgett, given the discussion we've had about automated
Twitter accounts and bots--and the range is obviously very
wide, but we know that's a problem. And you made an assertion
earlier that I want to come back to and just make sure it's
accurate. Do you require at Twitter, by service agreement, that
profiles are linked to real names, real people, or some other
way to make sure that those go back to real human beings, from
Social Security numbers to other unique identifiers?
Mr. Edgett. We do not. We require some information at sign-
up, but we don't require you to verify your identity. We have
services that verify identities on the platform.
Senator Heinrich. Why on Earth not?
Mr. Edgett. Because we see the power of Twitter being used
by folks like political dissidents, embedded journalists in
difficult countries who use the ability to not have to identify
themselves by name, like on other platforms, to speak their
truth to power. We see that----
Senator Heinrich. So the reason is for social dissidents
and people in third world countries or where there is a hostile
government regime? It is not your business model? You're not
reliant, for example, on those automated accounts to generate
revenue?
Mr. Edgett. We don't rely on--there's some good automation
on the platform and I'm happy to talk about that. But we do not
rely on this, the bad, malicious automation that we're talking
about here.
Senator Heinrich. If I were running a political campaign
today and I were to advertise on local television, on cable
television, in print or on the radio, or even through the mail,
I would have to have a ``paid for by'' disclaimer on those ads.
Now, Mr. Walker I believe has already addressed this issue. But
is there any policy reason that online social media ads, given
how effective and influential they have clearly become,
shouldn't meet that same level of transparency?
Mr. Edgett. We agree with the transparency efforts and last
week announced that we're creating a transparency center, not
just for political ads, which will have even more information
than all ads, but a transparency center for all ads, so that
you can see not just the ad that you've seen and why it's been
targeted to you, but all of the other ads created from that
same advertiser.
On the election front, you'll also be able to see who's
paying for the ad, how much they've spent on this ad campaign
and all ad campaigns, and you'll--able to see what the
targeting criteria are, so to better educate around why these
ads are on the platform.
Senator Heinrich. I appreciate that, Mr. Edgett.
Mr. Stretch.
Mr. Stretch. The same is true for Facebook. We are working
both on political ad transparency, enabling more visibility
into campaign ads by third parties, and also enabling campaigns
to meet their disclosure obligations in their--in their online
communications.
Senator Heinrich. Thank you, Mr. Chairman.
Chairman Burr. Senator Blunt.
Senator Blunt. Thank you, Mr. Chairman.
So, Mr. Edgett, in response to Mr. Heinrich's question,
there was a lot of information that you could get based on that
policy, if you pursue it, like all the other ads they ran, how
much. Would you get that by going to another spot? Surely
that's not all right there on the ad.
Mr. Edgett. Obviously we're a character-constrained
platform, so we will be identifying very clearly whether or not
something is a political ad, so that you can--you can see it
right away. And then, depending on if you're on a web browser
or on a mobile phone, you'll have to hover over or click on a
spot to then see a sort of a full transparency center that
gives you all that information right away.
Senator Blunt. So would you be able, on the ad itself, to
go ahead and put enough of a disclosure there that it's clear,
when you're looking at the ad, who paid for it and how to find
more information out about who paid for it?
Mr. Edgett. We're still working through the technical
details, but believe we'll be able to get that in front of----
Senator Blunt. Mr. Walker, are you trying to do anything
similar to that?
Mr. Walker. We are. Our idea is to have an icon that a user
can click on, so it's immediately available to them when they
see the ad.
Senator Blunt. But it wouldn't necessarily be--would any
information be on the ad except the icon?
Mr. Walker. It depends on the format of the ad. There may
be, in display context or video context, where it makes more
sense. In a very small amount of space, the FEC has struggled
with figuring out appropriate disclosure requirements. We
continue to look at that.
Senator Blunt. And you're looking at the other disclosure
requirements that other ads have to have on other media as
you're considering this?
Mr. Walker. Of course, broadcast, newspapers, and online.
Senator Blunt. Mr. Stretch.
Mr. Walker. Yes.
Mr. Stretch. The same is true for Facebook.
Senator Blunt. Looking at this same thing.
Well, on--when you're talking--and I think I'll start with
Mr. Edgett on this. When you're talking about Russians, are you
referring to the Russian government, or any Russian citizen, or
people who paid in rubles?
You mentioned the IRA, which I assume is not either your
individual retirement account or the Irish Republican Army. So
how do you know they're Russians? And what are you looking for
there when you're talking about Russians in retrospect?
Mr. Edgett. Right. That's a great question. We are looking
for signals. Not everyone identifies themselves as a Russian,
especially these malicious actors. So we're looking at things
like whether they registered in country in Russia. Do they have
a Russian phone number? Are they on a Russian mobile carrier?
Do they have a Russian email address? Are they coming in from a
Russian I.P.? Have they ever logged in--you'll see in our
retrospective work, we looked at have you ever logged in at any
time from Russia.
There are some technical--some technical challenges with
that. The trail sometimes goes cold at data centers, where
information is being processed. We saw about 14----
Senator Blunt. Okay. Looking back at what the so-called--
the Russians, however we're defining that, did during the
election, when you're saying the Russians paid for these ads,
these are ads paid for by the Russians because you've now gone
back and checked groups like the Internet Research Agency, and
you now know that's a Russian group?
Mr. Edgett. Well, on the advertising side we also have some
additional data around banking information, because folks are
paying for these ads. But we didn't link the IRA accounts to
advertising in the election. But what we did was found nine
advertisers, based on the signals I talked about and also
banking information. Largely it was Russia Today, who we have
since removed as advertisers from the platform.
Senator Blunt. Let's see if they got their money's worth.
Everybody here has been involved in one way or another in
buying advertising. I've always had some sense that in
advertising you pretty much got what you paid for.
Mr. Stretch, how much money did the Russians spend on ads
that we now look back as either disruptive or politically
intended? Is that $100,000? Is that----
Mr. Stretch. It's approximately $100,000.
Senator Blunt. I meant from your company.
Mr. Stretch. Yes, approximately $100,000.
Senator Blunt. How much of that did they pay before the
election?
Mr. Stretch. The----
Senator Blunt. I've seen the number 44,000. Is that right,
56 after, 44 before?
Mr. Stretch. The ad impressions ran 46 percent before the
election, the remainder after the election.
Senator Blunt. Forty six percent? Well, if I had a
consultant that was trying to impact an election and spent only
46 percent of the money before Election Day, I'd be pretty
upset about that, I think.
So they spent $46,000. How much did the Clinton and Trump
campaigns spend on Facebook during--I assume before the
election?
Mr. Stretch. Yes.
Senator Blunt. They were more organized than the other
group.
Mr. Stretch. Combined, approximately $81 million.
Senator Blunt. Eighty one million dollars. And before the
election?
Mr. Stretch. Yes.
Senator Blunt. So $81 million. I'm not a great
mathematician, but $46,000 out of $81 million, would that be
like five one-thousandths of one percent? It's something like
that.
Mr. Stretch. It's a--it's a small number by comparison,
certainly.
Senator Blunt. Very small number by comparison. So the fact
that we're talking about it today, it certainly seems like they
got their money's worth after the election, whether they got it
before or not. We're still talking about five one-thousandths
of one percent of the Facebook money spent. And that was just
by Clinton and Trump? Or was that all the presidential
candidates put together?
Mr. Stretch. No, those were the----
Senator Blunt. The $81 million?
Mr. Stretch. Those were the Clinton and Trump campaigns,
combined.
Senator Blunt. Well, probably got more attention here than
they did.
I know Ferguson, Missouri, and Baltimore were a couple of
big targets in a lot of these ad campaigns. Is there a way you
could do this to where you principally target viewers in the
St. Louis area for Ferguson or in Maryland, for Baltimore?
Mr. Stretch. It's important to distinguish between our ad
tools and the organic tools. Our ad--our ad tools do permit
geographic targeting of content. Approximately 25 percent of
the ads that we've identified and turned over to the Committee
were geographically targeted to a region smaller than the
United States. Most of them were targeted on a State basis.
Organic content--unpaid posts, if you will--are not
geographically targeted.
Senator Blunt. And some of those targeted dollars were
spent in states where the election turned out not to be close
at all. Is that right?
Mr. Stretch. That's correct.
Senator Blunt. The other questions that we'll get to, maybe
in--maybe in writing later, but this--on the free media, I
think we have to be very thoughtful here about who decides
what's voter suppression and what's not; who decides what level
of speech is acceptable and what's not. It's an unbelievable
obligation that the government's never been very good at, and
an unbelievable obligation that it sounds like to me your
companies are all being asked to assume. And that'll be an
ongoing discussion, I think, of whether that's possible or not,
and the questions and problems that arise when somebody does
begin to decide what's acceptable to talk about and what's not,
and what discourages voters and what doesn't.
I would think the general election process these days would
discourage voters from participating, so maybe that would just
mean none of it could be discussed. But we will see how that
goes.
Thank you, Mr. Chairman.
Chairman Burr. Senator King.
Senator King. Thank you, Mr. Chairman.
Gentlemen, you've done a good job this morning. I must say,
though, I'm disappointed that you're here and not your CEOs,
because we're talking about policy and the policies of the
companies. And it's fine to send general counsel, but I think,
if you could take a message back from this Committee, if we go
through this exercise again, we would appreciate seeing the top
people who are actually making the decision.
I want to begin with two quotes, and I generally don't read
quotes, but these are so apt. The first one says this:
``Nothing was more to be desired than that every practicable
obstacle should be opposed to cabal, intrigue, and corruption.
These most deadly adversaries of republican government might
naturally have been expected to make their approaches from more
than one quarter, but chiefly from the desire of foreign powers
to gain an improper ascendant in our councils.'' That's
Alexander Hamilton in the 68th Federalist. He saw this coming.
The other is a more recent quote from a fellow named
Vladimir Kvachkov, who's a former GRU officer. And he said: ``A
new type of war has emerged in which armed warfare has given up
its decisive place in the achievement of the military and
political objectives of war to another kind of warfare:
information warfare.'' And that is exactly what we're talking
about here today.
I appreciate the Chair and the Vice Chair giving us the
context of what we're doing. Their visual demonstrations I
think were very vivid. And the warfare is the division of our
society. And it's not only us; it's the entire West.
We know that the Russians were involved in the French
election. We know that they were involved in the German
elections. We are now learning that they're involved in the
separation of Spain, and my understanding is they've set up
shop in Scotland, which is talking about an independence vote
from Great Britain. So this is a sophisticated, worldwide
strategy that applied here in 2016.
There is one other piece I'd like to add to what the Chair
and the Vice Chair did, and that is that it's still happening.
This is a service of the German Marshall Fund called,
interestingly, ``Hamilton 68,'' that follows hashtags on a
daily basis. And I just picked a day in September to show.
These are the hashtags that are being propagated, not created
by Russians, but these are 600 Russian websites that are using
these are the hashtags on these particular days.
[The material referred to follows:]
[GRAPHIC] [TIFF OMITTED] T7398.139
Senator King. The interesting thing, Syria's up there;
clearly, Russia has an interest there. But then we have the
NFL; and then we have ``boycott the NFL''; then we have ``Stand
for our anthem''; we have, ``Make America great again'';
Russia, ``Take a knee.'' In other words, they were tweeting on
both sides of the NFL dispute in order to exacerbate the
divisions.
One witness said to this Committee that their strategy is
to take a crack in our society and turn it into a chasm. And
that's exactly what we saw in 2016. My point here is it hasn't
stopped and it won't stop.
So we have to figure out what to do about it. And it seems
to me that there are three possibilities, one of which you can
make a significant contribution to; the other two, frankly, are
up to us.
The first is a technical defense, the kind of thing you've
already been talking about today: checking identities,
identifying the source of this kind of information. I want to
pursue that in a minute.
The second is we as a society have to understand when we're
being conned. I spent some time a year ago in Eastern Europe,
before our election. And the Eastern European politicians, all
they wanted to talk about was Russian meddling in their
elections. And I said, ``How do you defend yourself? You can't
undo the internet. You can't turn off the TV.''
They said, ``All of our people now get it, that this is
what the Russians are doing. And when they see one of these
postings, they say, `Oh, it's just the Russians again.' ''
We have to develop that level of sophistication so that we
know when we're being misled. And to me, this is the mentality
we all apply at the checkout counter of the supermarket and we
see a tabloid that says a movie star had a two-headed baby. We
say, ``Oh, that's just a tabloid.'' We need to apply that same
kind of sensibility to these kinds of fake news, misleading and
purposeful distortions.
The third thing that we have to determine, I think, is that
this country has to have some kind of cyber warfare deterrent
capacity. Right now, there's no price to be paid for meddling
in our democracy; and our adversaries have to understand that,
if they're going to undertake a campaign like this, there will
be a price to be paid; there will be results; and if they do X,
we are going to do Y to them. Right now, that doesn't exist,
and all of what the Russians did last year has basically been a
free pass. And I think that's a very difficult problem.
Now, let me ask the technical question. Mr. Stretch, can
you guys--could you, for example, require a date line on a
posting that said where it comes from, just like a news story
says ``Moscow, September 23rd''? Is there some way to identify
the source of information as it comes across your news feeds?
Mr. Stretch. Senator, it's a great--it's a great question.
We do permit users to identify the geographic location of the
post. We don't require it. There are oftentimes privacy
considerations that would prevent a user from----
Senator King. You could require it by country, couldn't
you? That's----
Mr. Stretch. There are many uses of our services, Senator,
where requiring people to designate their physical location
could be problematic.
I would make two other points. One is, because of the way
the internet is architected, your geographic location is--can
be disguised. That's something we need to work on in order to
make sure we're not being fooled, because I think your larger
point is an excellent one. The geography of the location of the
user, paired with the content they're serving----
Senator King. It's part of the information.
Mr. Stretch [continuing]. Is part of the information, and
we need to do a better job tuning our system to be more
sensitive to that.
Senator King. Mr. Walker, you said the people should know
what they're getting. When we get information, we know it's in
a--it's in a newspaper, we see the name of the author, we see
the date line. We're in a new information distribution world
here, and we need to think about how to apply some of the
principles that have helped us to assess that information. And
I hope that you all will continue to develop policies, just as
the newspaper business did 100 years ago, that help your
customers to analyze and assess the validity of the data. The
problem now is we're just taking what comes as it comes.
I'll end with this: I have a quote on my kitchen wall that
my wife found, and it says, ``The great problem with quotes on
the internet is determining whether they're authentic.''
Abraham Lincoln.
[Laughter.]
Thank you, Mr. Chairman.
Chairman Burr. Senator Lankford.
Senator Lankford. Gentlemen, thank you for being here. I
hope you hear loud and clear from this Committee there are lots
of questions. You're asking lots of questions. This is not an
opposition to free speech, though. This is actually a battle to
try to protect free speech.
We want to have good American dialogue, and the fear is
that your platforms are being used by foreign actors that want
to abuse our free speech. If two Americans have a disagreement,
let's have at it. Let's walk it through as two Americans. If an
outsider wants to be able to come do it, we do have a problem
with that. And we're trying to be able to work through that.
So we're grateful that you're here and to be able to walk
through this, and we look forward to cooperation together and
to be able to figure out how we actually resolve some of these
extremely complicated issues.
I do want to be able to push on a little bit this issue
about the type of ads and the type of content. Mr. Edgett, you
mentioned from your testimony that in Twitter, of the 131,000
tweets posted during the time--and I assume that means that
September to November time period that you were tracking--only
9 percent of those tweets from those Russia-targeted accounts
were actually election-related; the others were all social
engagement, other issues. And I think that's being lost in the
conversation, that only 9 percent of the tweets were election-
related.
Now, my question is, for all of your platforms, what are
you seeing from Russian-related accounts that you're tracking
now and trying to be able to pull down or identify? What are
the social issues that are being discussed right now from those
sites?
You can go back for the last six months if you want to. But
give me some examples of the type of social issues that they're
engaging with. One has already been mentioned by Senator King
and that's the NFL, either boycott the NFL or take a knee on
that. Has that been actively pursued on your sites? Give me
examples of topics here.
Mr. Stretch. So one example we saw following the election
was an effort by the accounts we've identified to inflame some
of the post-election demonstrations we saw. So some of the
accounts turned to questioning the Electoral College, as an
example.
Senator Lankford. Okay. So let me walk through multiple
examples here. So Electoral College is one of them. Some of
them, it's been publicly reported that the site was used to try
to organize events that were election protest events in certain
cities, that--was sending out messages saying, ``We're all
going to protest at this spot,'' but obviously it was created
by a Russian group.
What else? NFL? Yes or no on that?
Mr. Edgett. We've seen that activity.
Senator Lankford. Okay. What else? What other issues?
Mr. Walker. We've seen more limited use of our services,
Senator, but among that I would say police shootings and racial
issues.
Senator Lankford. Okay. What else?
Mr. Stretch. Certainly immigration has remained a topic
throughout.
Senator Lankford. Okay. Any other issues?
[No response.]
This is part of the reason that multiple members--and
Senator Heinrich mentioned it earlier--we really do want these
ads to get out in the public space. We think there's great
value for all of your platforms to be able to say, ``This is
the type of content that foreign actors are actually trying to
put out, that are divisive content. When we put it out, that's
one thing; when you put it out, it's completely different.''
We think there's a great benefit for you to be able to say,
when you're aware of things, ``Please note, this is the type of
issue that's been coming out, that's coming out, and this is
what it looks like,'' and so people can say, ``That's the type
of thing I've seen before,'' or they can say, ``I've actually
liked on that before and didn't have any idea that that was
Russian-related.'' That'd be helpful.
Can I ask a question? When were you aware of Russian
activities on your platform during the election time? During
the election or before your election? When were you aware that
entities within the Russian government, whether that be the
Internet Research Agency or other individuals that you knew of
that were Russian, either government-related or at least
policy-related, were involved in election issues on your
platform? Before or after the election? And all three of you
can say. Before or after?
Mr. Stretch. We were aware of Russian state actors active
on the platform prior to and through the election, separate
from the Internet Research Agency. And we communicated with law
enforcement about our concerns at the time. These actors were
engaged in more traditional cyber threat activity, focusing on
account compromise, as well as trying to direct attention to
stolen information that was hosted on other sites.
Senator Lankford. How far back? Early 2016, 2015?
Mr. Stretch. We had seen activity as early as 2015.
Senator Lankford. Okay. Mr. Edgett.
Mr. Edgett. Twitter also saw activity from the IRA on our
platform and took large-scale action to take down accounts in
2015. We generally are aware of intelligence community reports,
so became aware of the activity in the report that came out in
January of this year; and then obviously, through the
retrospective of work, have uncovered what we think is the full
extent. We're continuing to look and to research that issue.
Senator Lankford. Okay. Mr. Walker.
Mr. Walker. We've been looking at cyber espionage issues,
accounts compromise issues, for many years. It was only after
the issuance of the report in January that we took a deeper
dive on the particular kinds of things we're talking about here
today.
Senator Lankford. One of the things that we're trying to
address is getting additional information. Getting the ads I
mentioned before would be very helpful. Getting the type of
posts that have been done would be very helpful to the American
people to be able to see.
Getting the statistical information would also be helpful.
The statistics that you've given us are the number of accounts
that are related to this, but it's not giving us the breadth
and the depth of those accounts once they're activated. So for
instance, it's helpful to know that one percent of those
accounts were Russian-related. But what we're not getting
statistically is was that one percent the first one percent?
They actually created it, launched it, and then millions of
people saw it after that? Are they in the middle, stirring it
and advancing it? Or were they beginning it?
That would be helpful to us as well. And I know you have
that data to be able to see if they started it or were they in
the middle of broadcasting it. So just giving us the percentage
of users that shared it doesn't help us. Knowing when in that
process and what happened after that is what helps us.
You all have done a lot of work on terrorism, on Islamic
extremism, on the advance of ISIS. You've done a lot of work on
child pornography, on human trafficking, on the sales of
illegal drugs on your sites. We're asking for help on this area
as well. And this is something that I would hope that we do not
have to engage legislatively or, if we do, it's the lightest
possible touch. This is something you have been actively
engaged in with other topics, and we look forward to
cooperation in this as well, so that you all are managing that.
But I do hope in the days ahead we continue to protect a
platform for free speech, to allow individuals to be able to
speak their opinions, whether its agreed or disagreed on by
other Americans, so Americans can engage in that dialogue.
Thank you.
Chairman Burr. Senator Manchin.
Senator Manchin. Thank you, Mr. Chairman.
Thank you all. RT, the Russian state news organization that
Federal intelligence officials call the Kremlin's principal
international propaganda outlet, RT America TV broadcasts
negative programs, derogatory information about the United
States. It's essentially information warfare directed against
the United States. According to the unclassified intelligence
report released on the 2016 elections, the Kremlin staffs RT
and closely supervises RT's coverage, recruiting people who can
convey Russian strategic messaging.
Additionally, the Kremlin spends $190 million a year on the
distribution and dissemination of RT programming. Can you all
answer, do any of your organizations allow RT to purchase space
or advertise on your accounts?
Mr. Edgett. We've off-boarded, which means we've banned,
Russia Today and their related accounts from advertising on the
Twitter platform.
Senator Manchin. That's effective as of when?
Mr. Edgett. That was effective as of a week or so ago.
Mr. Walker. We have reviewed the RT accounts. As you know,
RT advertises extensively in newspapers, magazines, airports
across the United States. We've not found violations of our
accounts. But we are focused on making sure that there is
enhanced transparency with regard to government-funded
broadcasting of all kinds, whether that's RT or Al Jazeera or
other sources.
Senator Manchin. Mr. Stretch.
Mr. Stretch. The same is true for Facebook, Senator.
Senator Manchin. Well, it says here that RT uses Google's
YouTube, Facebook, and Twitter as the main distributors of
their content. So you all have been the main go-to to get their
propaganda out to the United States and do harm to us.
Mr. Walker. That may be true online, Senator, but of
course, RT is covered by--RT channels are included on major
cable television networks, satellite television networks, hotel
television choices, et cetera. So it is a problem that goes
beyond the web.
Senator Manchin. Let me ask you this question, if I may.
Kaspersky Labs, their antivirus software, do any of you all
have it on your personal computers? Your personal computers,
are you using Kaspersky Labs?
Mr. Walker. No, sir.
Senator Manchin. Does your company use Kaspersky?
Mr. Walker. I do not believe so.
Senator Manchin. Could you check and find out?
Mr. Walker. We will--we will follow up.
Senator Manchin. How about any of the other two?
Mr. Edgett. Not aware, but we'll follow up with your staff.
Mr. Stretch. Likewise.
Senator Manchin. Do you all have any doubts about Russia's
involvement to interfere in our 2016 presidential election? Do
you have any doubts at all that Russia intervened, interfered,
had a profound outcome, basically, on our outcome to our
election? Do you all have any thoughts on that? Do you have any
doubts about that?
Mr. Stretch. We have no doubt that there were attempted
efforts at interference. It's something we're focused on
addressing going forward. In terms of whether it had an effect
on the outcome, that's not something we're in a position to
judge.
Senator Manchin. Well, let me ask this: Are you or your
CEOs concerned about the threat and damage your companies can
do to the U.S. with your far-reaching power, and that you have
been identified as the major distributors of fake news? Are you
concerned about that? Do your CEOs talk about the threats to
the United States of America, where you're domiciled? Or is it
basically just a business model that you're worried about?
Mr. Edgett. We're deeply concerned. This is an issue that
we talk about constantly. As I said earlier, the first part of
this year, we pointed our entire engineering product and design
team on tackling the issues of information quality, abuse of
our systems and protecting our users.
Mr. Walker. I'd join that. It's absolutely a very serious
issue. The North Star of Google is to provide accurate,
comprehensive, relevant information to people. We don't always
get that right. But we have tens of thousands of engineers who
are trying to improve our algorithms, improve the rating
guidelines we use for individuals, address the problem of
breaking news, which is a very challenging one to get right
when there's just not a lot of content out there. So we take
this very seriously.
Senator Manchin. You all would agree, then, I guess, with
the legislation that's been introduced by different members and
bipartisan Members of this Committee and other committees, that
you all should be regulated and overseen the same as we do
other news medias?
Mr. Stretch, you have any idea, a comment? Are you all
going to fight back on that, lobby against it, or are you going
to basically support the legislation that's going to be needed
to make sure that the American people are getting the facts and
not fake news?
Mr. Edgett. Senator, we stand ready to work with the
Committee on legislation. Any particular pieces of legislation,
we're certainly happy to talk about.
Senator Manchin. You've seen the legislation that we've put
out there. You all three are lawyers. I'm sure you're watching
it very closely, you know, the legislation that we have. Do you
agree with the exemption? Would you support a change in law
that treats all political advertising on the internet the same
as print and broadcast ads, to require identification of
sponsors, basically the legislation we're putting forth?
Mr. Edgett. We're very supportive of the direction of the
Honest Ads Act, and have had really productive conversations--
--
Senator Manchin. You all we will be speaking in support of
these pieces of legislation?
Mr. Edgett. Yeah, we have some fine-tuning that we'd love
to talk about. But we, as you saw, put out our own transparency
center that very much aligns with the information that the
Honest Ads Act was asking us to provide to you.
Senator Manchin. According to an Oxford University study
released in October, Russian trolls are now targeting American
military personnel and veterans on Twitter and Facebook.
Allegedly, these trolls are pushing fake news and injecting
them into veterans groups and to active-duty personnel.
In fact, one fake Facebook page was highlighted in a Stars
& Stripes article from October the 18th. The page was called
``Vietnam Vets of America,'' and it had attracted a follower
network of nearly 200,000. The real veterans organization,
Vietnam Veterans of America, called it an impostor page and
another example of how military and veterans are being targeted
with disinformation.
A week after the original article was reported on October
18th, Facebook reportedly took down the site, ostensibly for
violating intellectual property of the real veterans service
organization.
So I would ask Facebook and Twitter, is the story accurate?
Did you all know about this? And are you seeing our military
and our veterans being targeted? And how come it took so long
to take it down?
Mr. Stretch. Senator, we're intensely proud of the use of
Facebook by our military. You asked----
Senator Manchin. We're not talking about that----
Mr. Stretch. Right.
Senator Manchin. We're talking about the people who are
targeting them.
Mr. Stretch. Yes. And we are very focused on making sure
that it remains an authentic experience for them. We receive
many reports of inauthentic behavior. We try to act on them
quickly. We're trying to improve our tools to detect it even
before it's reported to us. I'm not familiar with the
particular--the particular page you've described.
Senator Manchin. Stars and Stripes?
Mr. Stretch. Yes, but I----
Senator Manchin. I hope you will check into that.
Mr. Stretch. I will certainly check into it, Senator.
Senator Manchin. Let me just say this, as my time's
wrapping up. You can see this is not a Democrat or Republican
issue. This is an American issue that we're concerned about,
the security of our Nation.
We're getting hit from every way you possibly can imagine.
And you all are the largest, one of the largest distributors of
news. And there can be no doubt that it has to be authentic and
true. You cannot allow what's going on against the United
States of America.
You are on the front lines with us, and we're doing
everything we can to support our military, our veterans, all
the people who put their lives on the line. And what you're
doing by allowing this fake stuff to come across, this
misleading, this damaging information, is really threatening
the security and the sovereignty of our nation.
I would hope that your CEOs--and I agree with Senator King.
I wish that your CEOs would be here. They need to answer for
this. It can't be a business model. It's got to be a security
issue.
Thank you.
Chairman Burr. Senator Cotton.
Senator Cotton. Thank you, gentlemen, for your appearance
this morning.
Mr. Edgett, I want to discuss Twitter's history of
cooperation with our intelligence community. Last year, in an
open hearing before this Committee, I asked then-CIA director
John Brennan about Twitter's decision to prohibit a subsidiary
called Data Miner from working with our intelligence community.
Director Brennan stated that he was disappointed in Twitter's
decision.
But at the same time that we learned that Twitter was
refusing to work with the CIA and the rest of the intelligence
community, we also learned that Twitter was pitching Russia
Today and Sputnik, propaganda arms of the Kremlin, to sell
advertisements for profit.
So, in essence, last year Russia was beginning its covert
influence campaign against the United States, and Twitter was
on the side of Russia, as opposed to the national security
interests of the United States. How can your company justify
this pattern of behavior to its fellow citizens?
Mr. Edgett. We work frequently and hard with law
enforcement all the time. We do have global policies that
prohibit the use of our data hoses or publicly available data
around tweets for purposes of surveillance. But we allow law
enforcement to use Data Miner and Twitter products around news
alerts, first response technology to see what's going on in an
area if a 911 call is made and an emergency responder is going
somewhere. But we do not allow surveillance based on Twitter
data.
Senator Cotton. Did Twitter cut off the CIA and the
intelligence community from Data Miner last year?
Mr. Edgett. We asked that our policy of surveillance be
applied consistently to all organizations, and I believe that
Data Miner has been enforcing that policy.
Senator Cotton. Has it cut off RT and Sputnik?
Mr. Edgett. As to Russia Today, when we approached Russia
Today last year to talk about our advertising products and to
sell them our advertising services, they were approached as a
regular media organization like a BBC or an NPR.
Senator Cotton. Do you consider RT to be a regular media
organization?
Mr. Edgett. Obviously not now. Coming out of the DNI report
earlier this year and the retrospective work that we've done
most recently, we don't. And that's why we have banned Russia
Today from advertising on the Twitter platform.
Senator Cotton. So there's a difference, though, between
the advertising question, which was improvident, and the use of
Data Miner. According to a Wall Street Journal report to which
Director Brennan was responding, Twitter CEO Jack Dorsey vetoed
the Data Miner-CIA contract at the last minute because he
objected to the, quote, ``optics'' of continuing to help U.S.
intelligence agencies.
That Wall Street Journal report also said, though, that
customers still getting Data Miner include RT. Is that an
accurate report? John Brennan had no reason to doubt its
accuracy.
Mr. Edgett. I'm--I don't--I don't have the information, but
will follow up on Russia Today's use of Data Miner's products,
which is the third party where we have a relationship. I
believe Mr. Dorsey wanted to make sure that our policies were
being applied consistently around surveillance.
Senator Cotton. Do you see an equivalency between the
Central Intelligence Agency and the Russian intelligence
services?
Mr. Edgett. We're not offering our service for surveillance
to any government.
Senator Cotton. So you will apply the same policy to our
intelligence community that you apply to an adversary's
intelligence services?
Mr. Edgett. As a global company, we have to apply our
policies consistently.
Senator Cotton. This reminds me of the old line from the
Cold War of one who did not see a distinction between the CIA
and the KGB on the other hand, because the KGB officer pushed
an old lady in front of an oncoming bus, the CIA officer pushed
the old lady out from the path of the oncoming bus, because
they both go around pushing old ladies.
I hope that Twitter will reconsider its policies when it's
dealing with friendly intelligence services in countries like
the United States and the U.K. as opposed to adversarial
countries like Russia and China.
Would Twitter entertain the possibility of once again
allowing the intelligence community to use Data Miner?
Mr. Edgett. We do today, for purposes of news alerts and
first response technology, getting information on certain
areas. We do not allow anyone--our policy is not to allow
anyone, for the purposes of user privacy, to use our technology
to run surveillance.
Senator Cotton. Okay, let's move on to another hostile
intelligence service. Other than Vladimir Putin and Russia, I
can't think of anyone who was more involved in efforts to
influence our election last year than Julian Assange and
WikiLeaks. The current director of the CIA, Mike Pompeo, as
well as this Committee in our annual Intelligence Authorization
Act, has labeled WikiLeaks a non-state hostile intelligence
service who aids hostile foreign powers like the Kremlin. Yet,
to my knowledge, Twitter still allows them to operate
uninhibited. Is that accurate?
Mr. Edgett. We have terms of service and rules that apply
to all users and will apply those consistently and without
bias. We take action on accounts like WikiLeaks----
Senator Cotton. Is it bias to side with America over our
adversaries?
Mr. Edgett. We're trying to be unbiased around the world.
We're obviously an American company and care deeply about the
issues that we're talking about today. But as it relates to
WikiLeaks or other accounts like it, we make sure that they're
in compliance with our policies, just like every other account,
and have and will continue, if we need to take------
Senator Cotton. So you'll be unbiased towards WikiLeaks and
Julian Assange, but you'll take down videos of people like
Marsha Blackburn, a Republican running for the United States
Senate?
Mr. Edgett. Marsha Blackburn's video was never removed from
the Twitter platform. She ran that tweet and that video as an
advertisement, and we have different standards for our
advertisements than we do for the organic tweets and content on
our platform, because we're serving ads to users who haven't
asked to follow Representative Blackburn or others and we want
to make sure that that's a positive experience. And so our
policies have a different standard. And in that case, we had
users reporting that it was inflammatory and upsetting, and it
was initially taken down.
We're making these tough calls all of the time, and in that
case we reversed the decision and allowed the advertisement to
continue to run. But we never took down Representative
Blackburn's tweet or allowed her not to convey that message to
those who were following her and to engage in the dialogue with
her.
Senator Cotton. Mr. Edgett, I know that you're the acting
general counsel, not the general counsel. And if you were the
general counsel, these decisions are made at the CEO and the
board of directors level. But I have to say most American
citizens would expect American companies to be willing to put
the interests of our country above, not on par with, our
adversaries--countries like Russia and China, or non-state
actors like WikiLeaks, or individuals like Julian Assange.
As many other Members of this Committee have expressed, I
think your companies have accomplished amazing things for our
country and its citizens, and made our lives better in many
ways. I also support the channels that you've created for free
speech, especially for some oppressed or persecuted people
around the world.
But this kind of attitude I would submit is not acceptable
to the large majority of Americans, and it's going to be part
of what would lead to unwise or imprudent regulation, not
sensible and smart regulation.
My time's expired.
Chairman Burr. Senator Harris.
Senator Harris. Thank you.
California is home to many of the world's most successful
technology companies and we're proud of that. And we also know
that with that great success comes great responsibility. Your
companies, therefore, have a great responsibility to the
American public.
And, as you know, you are the modern town square and the
modern postmaster. You are the phone company and the Yellow
Pages. You are the newspaper and the radio broadcaster and the
television station, and you are the emergency alert system.
Your decisions fundamentally inform public discourse.
So our Nation's enemies have used your platforms in a way
that has been designed to create and disseminate and advertise
hateful rhetoric, with the intent and the effect of disrupting
our democracy. And that, of course, is why we're all here.
I have several questions, but I'd like to start with what I
think is, frankly, an elephant in the room. And I'm holding up
the SEC Form 10-Q that has been filled out by each of your
companies. And your response to this is pretty much the same,
but I have in front of me that from Facebook. But Twitter and
Alphabet have the same information.
There's a section here which requests, ``What are the risks
related to the business?'' And it reads, ``If we fail to retain
existing users or add new users, that will be a problem for
us.'' It goes on to say, ``We generate substantially all of our
revenue from advertising. The loss of marketers or reduction in
spending by marketers could seriously harm our business.''
It goes on to say, ``Our advertising revenue could also be
adversely affected by a number of other factors, including
adverse legal developments relating to advertising, including
legislative and regulatory developments, and developments in
litigation.''
So my question to you is about American ads, not the
Russian ads, American ads that run on your platforms. There are
legitimate ads that appeared alongside of the Russian placement
and propaganda pages on Facebook, on Twitter and even on
YouTube.
So can you please tell me that, as it relates to those
advertisements on Facebook, on Twitter or in YouTube, how you
are addressing that, and in particular, how much money did you
make off of the legitimate ads that ran alongside the Russia
propaganda?
And we can start with Alphabet or with Google, please. And
that would be the advertisements that ran before your videos on
YouTube.
Mr. Walker. Sure. The total amount of advertising we
discovered across our platforms was $4,700 from the Russian
sources.
Senator Harris. That's not my question.
Mr. Walker. I understand.
Senator Harris. My question is American advertising or
legitimate advertising. How much money did you make from
legitimate advertising that ran alongside the Russia
propaganda?
Mr. Walker. A de minimis amount, Senator. I don't have it
in front of me. We'd be happy to follow up.
Senator Harris. Okay. What about for Twitter?
Mr. Edgett. I don't have the data, but I will follow up
and------
Senator Harris. Have you not looked into that?
Mr. Edgett. I believe--are you asking how much advertising
revenue we made for the period, totally? Or----
Senator Harris. I'm asking how much advertising revenue did
you receive from legitimate advertisers that advertised
alongside or in connection with Russian propaganda?
Mr. Edgett. We haven't done that analysis, but we'll follow
up and work on that.
Senator Harris. Okay. What about Facebook?
Mr. Stretch. The same is true for Facebook, Senator.
Senator Harris. You've not done that calculation?
Mr. Stretch. We've not done that analysis.
Senator Harris. I find that difficult to understand,
because it would seem to me that we would figure out how much
you've profited from Russian propaganda on your platforms. So
please do follow up with us as soon as possible on that.
And also, it is critically important that this Committee
have access to all of the information it needs to understand
the Russia propaganda in the 2016 election. So will you each
commit to retaining records, as you are required to do from the
minimum standard of media rating research, but do that and
extend it beyond the 11 months that they require and extend it
through the completion of our investigation into Russia's
interference in the 2016 election? Will you commit to keeping
those records during the duration of our investigation?
Mr. Walker. We will commit--we will keep all relevant
records to this investigation and provide them to the
Committee, yes.
Mr. Edgett. Same goes for Twitter.
Senator Harris. Thank you.
Mr. Walker. Yes.
Senator Harris. And as for all three of you, can you please
name the senior executive who is responsible in your operation
for countering state-sponsored information operations? And if
you do not have one, please indicate that, as well.
Mr. Walker. It's a challenging question, because we have a
number of people across different teams, including our cyber
espionage teams, as well as our trust and safety teams. I would
say our chief security officer is one such person. Another
person would be the head of our trust and safety team. Then we
also have separate teams at YouTube who work in these areas.
Senator Harris. So I take it you have not designated an
individual as part of your executive team who's responsible
specifically for state-influence operations?
Mr. Walker. I will take responsibility for that, Senator.
Senator Harris. Okay, I appreciate that.
Mr. Edgett. That role--there's two people filling that role
at Twitter. The first is our general counsel. I'm currently our
acting general counsel, so currently it is me. But also our
head of our Twitter product, the Twitter product that we all
use, has taken responsibility for safety, abuse, and
information quality on the platform. So I feel like that's
directly related to your question.
Senator Harris. Okay. But I'd like you each to appreciate
and everyone to appreciate that this is a very specific issue
with its own pathology, requiring a great amount of resources,
because we are talking about state-sponsored activity. This is
not about an individual conducting this activity and then you
need to review it.
So as it relates to state-sponsored information operations,
I'm requesting that you name whoever is responsible now, but as
we go forward that you designate in your operation someone at
the executive level who is responsible specifically for those
types of operations, understanding that, as we know now, there
are governments that are willing to put incredible amount of
resources into manipulating the American public.
And it is beyond what you might need to review in terms of
activity on your sites that involves issues of posting
inappropriate images and things of that nature.
Mr. Stretch.
Mr. Stretch. Senator, we have a chief security officer and
a threat intelligence team that's acutely focused on this
threat. I will take responsibility for our overall response to
this threat.
Senator Harris. And how many of your employees are
dedicated to addressing state-sponsored operations
specifically? And if there aren't, please follow up in terms of
what you're prepared to dedicate to that.
Mr. Stretch. This is a--it's a harder question, because
there are so many vectors that we're investing in. I stated
earlier that we have 10,000 people at Facebook across a number
of teams who are focused on safety and security generally, and
we're doubling that number. The number of people who think of
this as their full-time job is something I'll have to come back
to you on.
Senator Harris. Okay, I appreciate that. And for each of
the companies, we'd appreciate that.
You can create automated systems that detect foreign
propaganda. For example, you can determine whether a user is
active during Moscow business hours, or connects through a VPN,
or registers with a fishy voice over an IP telephone number.
And you can feed those signals into a machine that can actually
create an algorithm that can allow us to indicate or figure out
if propaganda is actually being pushed through. Have you done
that as it relates to state-sponsored manipulation of
elections?
Mr. Edgett. So our technology is agnostic. We have the
technology you're talking about, which is an algorithm that
helps us catch the bad actors based on their pattern and
behavior, and also connect accounts so that if they start new
accounts or new networks of accounts, we get those before they
tweet.
We want to catch that activity all over Twitter. That,
having automated accounts, malicious actors on Twitter, is a
bad experience for our users. So we've been tackling that
problem for years, and the challenge is as we get better, these
actors get better. And so it's a constant game of cat and mouse
and one-upmanship.
But we are committed every single day to making sure that
we are removing those actors from our platform.
Senator Harris. My time is running out, so perhaps we can
just have quick answers for the remaining folks.
Mr. Walker. Our answer would be similar.
Mr. Stretch. The same.
Senator Harris. Thank you.
Chairman Burr. Members should be aware that there has been
a series of votes started about seven minutes ago, two votes.
We've got two Members left. We're going to move through those
and wrap up.
And at this time, I would ask unanimous consent that all
Members be allowed for seven days to submit questions to our
witnesses today.
[No response.]
Without objection, so ordered.
Senator Cornyn.
Senator Cornyn. Thank you for being here.
It strikes me that the United States is operating at a
tremendous disadvantage. We are a free and open society. We
believe in freedom of the press, freedom of expression, and we
respect the privacy rights of individual--of individuals. Our
opponents have the opposite view. They view information as a
tool of warfare, while denying their own citizens access to the
sort of freedoms--the press, individual thought, and
expression--that we celebrate in this country. But it seems to
me that you're at the intersection of this problem, which
brings us to why you're here today.
I'll restate the Chairman's comments when he started by
quoting H.L. Mencken who said, ``For every complex problem,
there is a clear, simple, and wrong answer.'' And so we need to
be very careful, I think, in how we deal with this.
But I do think the public needs to understand how your
platforms operate. My friend Senator Harris talked about your
being the town square and the newspaper, the radio station. But
you are more than just a publisher of information. As you point
out, most of the income that your companies earn is from data
mining. You know more about individual Americans than anybody
else, including their own government. And, of course, you vow
to protect that information and their privacy at the same time.
But you use it in order to target ads. Many of us here on
the panel have used your platforms in political campaigns to
make sure that our message gets to voters who we think might be
receptive to our point of view or our platform.
But I'd like to know as a general matter how do you
distinguish between somebody, like Senator Feinstein alluded
to, using social media as a means to recruit and incite lone-
wolf terrorists like the one that plowed down unsuspecting New
Yorkers yesterday, killing 8 and injuring 11 others, between
that person and a foreign government using your platforms in
order to pursue the sort of disinformation and active measures
that caused so much confusion and polarization in our election?
And how do you distinguish between the way you treat those
people on your platforms and how you treat sex traffickers, who
are targeting young girls and selling them, in essence, for
sex? How do you distinguish between those three?
Mr. Walker. Senator, I'd say none of those activities are
acceptable on our platforms. We have strong policies against
abuse of platforms to promote hate, hate speech, incitement of
violence, sex trafficking, human trafficking, et cetera. In
addition, when you have the added layer of a deceptive actor,
like a foreign government, trying to push its points of view in
deceptive ways, without identifying itself, etcetera, that
separately violates another layer of our policies.
Mr. Edgett. It goes back to what we talked about earlier.
We also have the same policies prohibiting all of this. We've
had the best success so far looking at the behavior and the
signals we see only on the Twitter side, around how accounts
are linked and the activity of those accounts, where they're
logging in from, what they're doing with each other, to stop
these bad actors before they're even able to tweet.
A good example of this is on the terrorism side. We are now
able to automatically detect and take down 95 percent of
terrorist accounts, 75 percent of those before they tweet for
the first time. So because we're focusing on behavior and not
the content, we don't have to wait for them to get their
messages out. We take those down proactively before they're
able to speak.
Mr. Stretch. I would just add, Senator, that of the three
forms of abuse of the platform you describe--child safety,
terrorism, and foreign interference in elections--the first two
we've been working on for quite a long time as an industry and
with government, and I think we have a proven track record of
working well together, as an industry and with government, to
make sure we're taking appropriate steps to address those
abuses. We have work to do, but we have made progress.
This threat of foreign interference in the elections is
something where we do need to up our game, I think, as a
company and as an industry and working with government. But the
success we've made or the successes we've had on those first
two forms of abuse gives me some reason for optimism going
forward, with respect to our ability to address the foreign
interference threat.
Senator Cornyn. For each of you, have the terms of service
of your company changed at all since the 2015, 2016 actions of
Russia? Your terms of service?
Mr. Walker. Yes. We have changed our policies with regard
to ads that are appropriate, as well as content that's
available on YouTube.
Senator Cornyn. Focused on active measures or on other
matters?
Mr. Walker. Our terms have to do with terms around
advertising and terms about what acceptable content is. We are
simultaneously using additional tools behind the scenes to
identify material.
Mr. Edgett. It's very similar at Twitter.
Mr. Stretch. The same is true for Facebook.
Senator Cornyn. Why should your companies be treated any
different than the press from a legal accountability
standpoint?
Mr. Edgett. We believe, as a user-generated content
platform, that the rules around Section 230 provide a platform
to our users around free speech and expression, and don't
require us to take a bias on removing content that we fear will
violate certain rights. And so we work actively to prohibit
things like violence and terrorism and abuse and harassment.
And you'll see how we were tackling this problem with urgency
and seriousness.
But we believe that, as a user-generated content platform,
we want to allow the free expression and debate without the
interference of some of the things you're talking about.
Senator Cornyn. So you believe you should be treated from a
legal standpoint differently than a newspaper, cable TV show,
or a radio show?
Mr. Edgett. Yes. We're not producing the content; we're
allowing users to upload. We have a lot of great journalists
and news organizations who are putting content on our platform
to share, linking back to their sites. We're offering the
service to allow that interchange, that information sharing.
Senator Cornyn. That may well be a distinction that is lost
on most of us, that you're just a platform for other people to
express their views, as opposed to being a publisher in your
own right of those, of those views.
Finally, let me just ask each of you, please, to continue
to work with us on the Stop Enabling Sex Traffickers Act of
2017. As you know, this deals with the Communications Decency
Act, which has been used as a legal bar to those who have been
victims of sexual abuse, when they seek to bring the people who
facilitated that sex trafficking to justice. And I think there
is a way that, working together, we can come up with something
that protects the victims, but also maintains the freedom of
the internet. And I would just encourage each of you to
continue working with us on that, so we can reach an acceptable
outcome.
Thank you, Mr. Chairman.
Chairman Burr. Senator Reed.
Senator Reed. Thank you, Mr. Chairman.
For all the panelists, but starting with Mr. Stretch: When
you discover a deceptive foreign government presentation on
your platform, my presumption is from what you've said today
you'll stop it and take it down. Do you feel an obligation in
turn to notify those people who have accessed that? And can you
do that? And shouldn't you do that?
Mr. Stretch. Senator, we feel an obligation, as you say,
first to stop the activity; second, to investigate it further,
to fan out, essentially, from the account to make sure we're
taking an expansive view of the investigation to try to capture
any related activity; third, to share threat information with
the industry and with the government so we can all do a better
job; and then, fourth, to bring the issue to the attention of,
in this case, this Committee.
And the content itself, we've said we're supportive of this
Committee making it publicly available. The question of
reaching out to individuals who may have seen it is a much more
difficult and complex one. But we believe our commitment to
transparency on this issue generally should address that.
Senator Reed. Well, potentially you could do that, I
presume, or you could invest the resources to do it. And as a
result, frankly, reporting to us about the nefarious activities
of Russia is not going to immediately translate to the
thousands or apparently 126 million people who saw the message
and thought it was legit.
You have the, I presume, the technical skill to do that.
Again, apropos Senator Cornyn, you know, you'll see in the
newspaper, ``We correct the statement we made the other day; it
was wrong,'' or ``It was deliberately wrong.'' And I think you,
given the First Amendment, you can live with that, I hope.
Mr. Stretch. I'm sorry, Senator. Could you repeat the
question?
Senator Reed. Well, the question goes back to having an
obligation under the First Amendment to notify people who you
know have been deliberately misled by a foreign government--not
just us, not just law enforcement.
Mr. Stretch. The technical challenges associated with that
undertaking are substantial, particularly because much of the
data work underneath our estimate of the number of people who
may have been exposed to this relies on data analysis and
modeling.
That said, we do believe transparency in this area is
important, and we are supportive of making as much of this
information available to the public as the Committee deems
warranted.
Senator Reed. Mr. Edgett. I deem it warranted, for whatever
that's worth.
Mr. Edgett.
Mr. Edgett. It's an interesting proposition. We have a team
dedicated to information quality and how we present information
on the platform. We see, as an open platform, active dialogue
around a lot of this false information, fake information right
away. So when you're seeing the tweets, you're also seeing a
number of replies to it, showing people where to go, where
other information is that's accurate. But we will definitely
take that idea back to explore how we could implement a process
like that.
Senator Reed. Mr. Walker, your platform?
Mr. Walker. We're somewhat differently positioned in that,
because were not primarily a social network, many of our users
are not logged in at the time they access content. So it's
difficult for us to know exactly who has seen what. But we too
will take it under consideration.
Senator Reed. Thank you.
With respect to bots, what I've gleaned from the testimony
is that you can technically identify a bot system operating on
your platform; and then, am I right to assume that you will
shut them down? Any bot system, you will shut down?
Particularly a government-related bot system?
Mr. Walker. When we refer to bots, it's primarily a Twitter
issue. I'm not familiar with bots on Google per se.
Senator Reed. Well, let's go to Twitter. I'm not the
technical expert, but YouTube is your subsidiary, I presume.
Mr. Walker. That's correct, yes.
Senator Reed. And a lot of the hits on YouTube I presume
were generated by electronic devices, not people, so that the
RT program that was attacking Secretary of State Clinton had a
2 percent over-the-air audience, but a huge number of hits on
YouTube. And as a result, you know, you are being unwittingly
or wittingly used by bot systems all of the time.
Mr. Walker. A bot, an automation, an automated view of
content, really isn't sort of the core issue. What people tend
to do on YouTube is try to drive up their perceived view counts
to make themselves appear to be more popular than they are.
Senator Reed. And they do it by using electronic networks.
Mr. Walker. That's certainly right. And so we--yes,
across--and this is a problem not limited to this context. Many
people would like to make themselves appear more popular than
they are. So we have a lot of sophisticated tools that are
precisely designed to combat that kind of phenomenon, yes.
Senator Reed. And you will, if you find it, reduce the
number of hits so it's no longer trending?
Mr. Walker. Either that or remove people from our services
for abuses of our terms of service, yes.
Senator Reed. So you do have to deal with these bot
networks, and you're dealing with it.
Mr. Walker. In that sense, that's correct, yes.
Senator Reed. Okay.
Bots?
Mr. Edgett. Similarly, we remove as many bad automated
accounts as we can find. As I said earlier, some of these bad
actors are trying to get more sophisticated, and so we're
staying ahead of that by learning from the automated accounts
that we're seeing. But we will remove them, and have technology
to make sure that automated accounts aren't gaming trends, so
that--the trending hashtags that people are seeing on the
platform. We will remove their content from search and from the
timelines, and we will remove them permanently from the system,
you know, once we are able to investigate.
Senator Reed. Mr. Stretch, please.
Mr. Stretch. Senator, we--apologies.
We prohibit automated account creation, and in doing so
we're always looking for evidence of accounts being created en
masse and engaging in either of the behaviors that Mr. Walker
and Mr. Edgett identified.
Senator Reed. Thank you.
Just a quick question, because my time is going. This is a
daunting effort. We're being attacked, and you have to go on
the offense, the counter-offense, because the way we've
structured this system, we have very limited government role in
your regulation, your activities, et cetera.
And it comes down ultimately to resources. So I would like
to follow up officially. But what percent of your revenue are
you devoting to these activities? I mean proactive activities,
not, you know, just if someone complains enough we'll take it
down, but finding bots, thinking about notifying recipients of
clearly bad information. So what percent do you think, Mr.
Stretch, right now?
Mr. Stretch. Senator, I cannot give you a percentage. I can
tell you that the company is committed to getting this right.
Our most expensive resources are--resources are people and, as
I stated earlier, we are doubling the amount of people who will
be focused on these efforts in the coming year.
Senator Reed. Can you get us a number, please, in writing?
Mr. Stretch. Yes.
Senator Reed. Thank you.
Mr. Edgett, please.
Mr. Edgett. We also dedicate a lot of resources to this,
and I'll follow up with your staff.
Senator Reed. Mr. Walker, similarly?
Mr. Walker. Similarly.
Senator Reed. Thank you.
Thank you.
Chairman Burr. The Chair recognizes the Vice Chair.
Vice Chairman Warner. Thank you, Mr. Chairman. I'll be very
brief. I know our time is limited.
One, I want to acknowledge Senator Reed I think raised a
very good question. And if you were in a medical facility and
you got exposed to a disease, the medical facility would have
to tell the folks who were exposed. The comment as well about
TV and radio making corrections. I do think it's an interesting
question about what obligation you might have.
I just have to tell you, I think there has been some
progress made and I appreciate some of your efforts. I still
find it very disturbing that it appears that, at least from
Twitter and Facebook, the sense is that all the Russian active
measures only originated with one single troll farm in St.
Petersburg, and it still appears that most of the work that you
have provided us is derivative of your initial reports.
I was hoping very much that you would come in today and
either say that was absolutely all of it or we've identified
other troll farms or other entities. And I think we've got a
lot more work to do.
Thank you so much, Mr. Chairman, for this hearing.
Chairman Burr. Thank you. Thank you, Senator Warner.
We've come to the close of this and let me just make a
couple of statements, if I can. If, for some reason, you need
antitrust waivers to collaborate with each other, please let us
know. More importantly, seek for the waiver yourself. This is
going to take an overall effort to minimize--I'm not going to
use the term ``eliminate''--to minimize the damage and, more
importantly, the impact of what Russia is doing, did do, and
what others will do next year.
I firmly believe that all three of your companies have a
new perspective on security and that you've got your varying
degrees of changes that are good. The challenge is if it fails
the impact of that failure is significantly different than it
was in the 2016 elections. I need you to know that up front.
You all acknowledge that FEC law applies to you, but it
hasn't been lost on me that all of you asked for an exemption
from the applicable FEC law. So I sort of am reminded that a
portion of the content posted on Facebook by foreign actors
appeared to support one candidate or another. Clearly, it falls
within the lines of what the FEC law was there for. And, as you
know, Federal campaign law requires disclosures of sources for
ads. I have to put that little thing, ``Paid for by.''
Mr. Edgett, you said your guys are adopting something
similar. I applaud you on that. And I'm sure, if I asked why
you didn't apply it, you might have told me before that the FEC
law didn't apply to you because they had a hung jury at the
FEC. They didn't decide it did, and they didn't decide it
didn't. Or maybe it's the excuse that it was small, impractical
items, therefore they had an exception to the disclosure.
Let me make it perfectly clear: There is no exception to
the disclosure as it relates to foreign money used to influence
U.S. elections. That is a national security issue. It is a
direct attempt to infiltrate the democracy that we have here.
And if it's not stated in the law, it should be your company's
responsibility to take it on head-on.
So in the future I hope that, if there's a takeaway from
this, it's that everybody's going to adhere to FEC law. If
you're a media outlet and you've known that it applied to you;
if for some reason you ever questioned whether it was foreign
money, then you probably didn't run the ad. I hope that none of
your platform is conditional upon you not verifying where that
money's coming from.
So the one thing that I didn't hear today, and I hope you
will take it back, is the effort to certify who's paying for
these things. I, like others, do not want the government to
stipulate to any of you what content, especially political
content, should look like. By the same token, you're the front
line before anything else has to kick in to certify that
foreign money is not finding its way into influencing U.S.
elections.
I wish I could leave today and believe the only thing we
have to worry about are elections. The truth is we've spent a
tremendous amount of time as two old guys trying to figure out
what bots were and things that I don't use. But I have to
understand them in a way that as a policymaker makes decisions
that are best for the future of my kids and my grandchildren,
to make sure they've got access to your platforms, to make sure
that they can experience things I never dreamed about
experiencing and am too old to understand.
But I recognize the fact that I can't be influenced just
because I don't understand it, because I've got to match my
capital with the intellect of all of you at the table and the
people that work at all your companies. That is our future.
Don't let nation-states disrupt our future. And you're the
front line of defense for it. Please take that back to your
companies.
This hearing is adjourned.
[Whereupon, at 12:30 p.m., the hearing was adjourned.]
Supplemental Material
[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]
[all]