Hearings
Hearing Type:
Open
Date & Time:
Wednesday, September 5, 2018 - 9:30am
Location:
Dirksen G50
Witnesses
Sheryl
Sandberg
Chief Operating Officer
Facebook
Jack
Dorsey
Chief Executive Officer
Twitter
Larry
Page (Invited)
Chief Executive Officer
Alphabet
Full Transcript
[Senate Hearing 115-460]
[From the U.S. Government Publishing Office]
S. Hrg. 115-460
OPEN HEARING ON FOREIGN INFLUENCE
OPERATIONS' USE OF SOCIAL MEDIA PLATFORMS
(COMPANY WITNESSES)
=======================================================================
HEARING
BEFORE THE
SELECT COMMITTEE ON INTELLIGENCE
OF THE
UNITED STATES SENATE
ONE HUNDRED FIFTEENTH CONGRESS
SECOND SESSION
__________
WEDNESDAY, SEPTEMBER 5, 2018
__________
Printed for the use of the Select Committee on Intelligence
[GRAPHIC NOT AVAILABLE IN TIFF FORMAT]
Available via the World Wide Web: http://www.govinfo.gov
__________
U.S. GOVERNMENT PUBLISHING OFFICE
31-350 PDF WASHINGTON : 2019
-----------------------------------------------------------------------------------
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
JAMES INHOFE, Oklahoma, Ex Officio
JACK REED, Rhode Island, Ex Officio
----------
Chris Joyner, Staff Director
Michael Casey, Minority Staff Director
Kelsey Stroud Bailey, Chief Clerk
CONTENTS
----------
SEPTEMBER 5, 2018
OPENING STATEMENTS
Burr, Hon. Richard, Chairman, a U.S. Senator from North Carolina. 1
Warner, Mark R., Vice Chairman, a U.S. Senator from Virginia..... 3
WITNESSES
Sandberg, Sheryl, Chief Operating Officer, Facebook.............. 6
Prepared statement........................................... 9
Dorsey, Jack, Chief Executive Officer, Twitter, Inc.............. 19
Prepared statement........................................... 21
SUPPLEMENTAL MATERIAL
Responses to Questions for the Record by:
Sheryl Sandberg.............................................. 68
Jack Dorsey.................................................. 133
OPEN HEARING ON FOREIGN INFLUENCE
OPERATIONS' USE OF SOCIAL MEDIA
PLATFORMS (COMPANY WITNESSES)
----------
WEDNESDAY, SEPTEMBER 5, 2018
U.S. Senate,
Select Committee on Intelligence,
Washington, DC.
The Committee met, pursuant to notice, at 9:32 a.m., in
Room G-50, Dirksen Senate Office Building, Hon. Richard Burr
(Chairman of the Committee) presiding.
Present: Senators Burr, Warner, Risch, Rubio, Collins,
Blunt, Lankford, Cotton, 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. And
I'd like to welcome our witnesses today: Jack Dorsey, chief
executive officer at Twitter--Jack, welcome--and Sheryl
Sandberg, chief operating officer at Facebook. I thank both of
you for being here with us this morning.
Before I make my remarks, I want to say a few words about
our colleague, our friend, and committee ex officio member
Senator John McCain.
John could be blunt, and he could be direct, but when it
came to committing himself to a cause that he believed in, John
McCain was without equal. This Senate, this deliberative body,
with its history and its traditions, will survive the passing
of John McCain, but there can be no denying that the place is a
little smaller without him. We will continue to do the
important work we do here with passion, resolve, and a sense of
purpose born from moral conviction. John would want that. In
fact, he would insist on it from each of us.
My friends, if I can borrow the phrase: Arizona's loss is
our loss, and our loss is America's loss. John McCain will be
dearly missed, and as you can see, we have set his spot on the
dais today.
Jack, Sheryl--as a committee, we've learned more about
social media over the last 18 months than I suspect most of us
ever thought we would in a lifetime. We've learned about social
media's boundless potential for good and its ability to enable
thoughtful and engaged interactions on a global scale.
But we've also learned about how vulnerable social media is
to corruption and misuse. The very worst examples of this are
absolutely chilling and a threat to our democracy: the founding
ideal of different people from different beliefs and ideas all
living peacefully under a single flag. The committee takes this
issue very seriously and we appreciate the fact that Facebook
and Twitter are represented here this morning with an
equivalent and appropriate measure of seriousness.
The purpose of today's hearing is to discuss the role that
social media plays in the execution of foreign influence
operations. In the past, we've used terms like misinformation
and divisive content to describe this activity.
Now as we go into our fourth and final hearing on this
subject, I think it's important that we be precise and candid
with our language, because that's what the significance of this
threat demands. We need to be precise about the foreign actors
we're talking about, we need to be precise about the
consequences of not acting, and we need to be candid about
where responsibility for solving this problem lies.
Two weeks ago your companies announced a series of
successful disruptions that resulted in the removal of 652
Facebook pages, groups, and accounts, and 284 Twitter accounts
based on their violating your company's standards of
coordinated manipulation and inauthentic behavior. Google's own
internal security teams did commendable work disrupting this
influence operation and we would have valued the opportunity to
speak with them at the appropriate level of corporate
representation. Nevertheless, their efforts should be
acknowledged.
In a departure from what we've all gotten a little
accustomed to, this activity didn't come from Russia. It came
from Iran. My instinct is to applaud the diligence of your
security teams and credit you with taking the problem very
seriously.
But I'm not sure your success is the big story here. As I
understand it, a third-party security team was crucial to
identifying the scope of the Iranian activity. And even more
concerning is that more foreign countries are now trying to use
your products to shape and manipulate American political
sentiment as an instrument of statecraft.
Jack, I was pleased when informed about your efforts to
improve conversational health at Twitter. I think that kind of
initiative can do a lot to improve the transparency of public
discourse on your platform, and foreign influence operations
thrive without transparency.
Sheryl, I fully support Facebook's hiring of the right
security experts, building the necessary technologies and
collaborating across law enforcement, commercial,
cybersecurity, and social media company lines.
I think the observation that no one company can fight this
on their own is spot on. Unfortunately, what I described as a
national security vulnerability and an unacceptable risk back
in November remains unaddressed. That risk and vulnerability
was highlighted yet two weeks ago. Without question, positive
things are happening. The collaboration, dedication, and
resources and demonstrated willingness to work with us are
critical and valued by every member of this committee.
It takes courage to call out a state actor and your
companies have done that. But clearly this problem is not going
away. I'm not even sure it's trending in the right direction. I
will go back to what I said up front: we need to be candid
about responsibility, and by that, I mean both the
responsibility we have to one another--from one side of this
dais to the other--as participants in this public policy
discussion. And more importantly our shared responsibility to
the American people.
Technology always moves faster than regulation, and to be
frank, the products and services that enable social media don't
fit neatly into the consumer safety or regulatory constructs of
the past. The old definitions that used to differentiate a
content publisher from a content facilitator are just not
helpful here. I think that ambiguity has given rise to
something of a convenient identity crisis, whereby judgments
about what is and isn't allowable on social media are too
episodic, too reactive, and too unrestricted. People are
affected by the information your platforms channel to them.
That channeling isn't passive or random. It's a function of
brilliant algorithms and an incentive structure that prizes
engagement. None of that is under attack here.
What is under attack is the idea that business as usual is
good enough. The information your platform disseminates changes
minds and hardens opinions. It helps people make sense of the
world. When you control that or you influence a little of it,
you're in a position to win wars without firing a shot. That's
how serious this is.
We've identified the problem. Now it's time to identify the
solution. Sheryl and Jack, I'm glad you decided to appear and
your willingness to be part of the solution. I'm disappointed
Google decided against sending out the right senior-level
executive to participate in what I truly expect to be a
productive discussion.
If the answer is regulation, let's have an honest dialogue
about what that looks like. If the key is more resources or
legislation that facilitates information sharing and government
cooperation, let's get it out there. If it's national security
policies that punish the kind of information and influence
operations we're talking about this morning, to the point that
they aren't even considered in foreign capitals, then let's
acknowledge that. But whatever the answer is, we've got to do
this collaboratively and we've got to do it now. That's our
responsibility to the American people.
I'll offer a closing point. This is for the witnesses and
the members alike. There are no unsolvable problems. There is
only the will to do what needs to be done--or its absence.
With that, I turn to the Vice Chairman for any comments.
OPENING STATEMENT OF HON. MARK R. WARNER, VICE CHAIRMAN, A U.S.
SENATOR FROM VIRGINIA
Vice Chairman Warner. Thank you, Mr. Chairman. And let me
first of all echo your comments about our colleague and friend,
John McCain. I hope we all take his advice to continue to put
country first.
Welcome to the witnesses. Mr. Chairman has pointed out that
today is an important public discussion. I am pleased that both
Facebook and Twitter have sent their company's top leadership
to address some of the critical public policy challenges. I
look forward to a constructive engagement.
I'd say, though, that I am deeply disappointed that Google,
one of the most influential digital platforms in the world,
chose not to send its own top corporate leadership to engage
this committee. Because I know our members have a series of
difficult questions about structural vulnerabilities on a
number of Google's platforms that we will need answers for:
from Google Search, which continues to have problems surfacing
absurd conspiracies; to YouTube, where Russian-backed
disinformation agents promoted hundreds of divisive videos; to
Gmail, where state-sponsored operatives attempted countless
hacking attempts. Google has an immense responsibility in this
space.
Given its size and influence, I would have thought that
leadership at Google would have wanted to demonstrate how
seriously it takes these challenges and actually take a
leadership role in this important discussion. Unfortunately,
they didn't choose to make that decision. But for the two
companies that have chosen to constructively engage and to
publicly answer some difficult and challenging questions,
again, thank you.
Now, it would be an understatement to say that much has
changed in the aftermath of the 2016 campaign. With the benefit
of hindsight, it's obvious that serious mistakes were made by
both Facebook and Twitter. You, like the Federal Government,
were caught flat-footed by the brazen attacks on our election.
Even after the election, you were reluctant to admit there
was a problem. I think in many ways it was pressure that was
brought to bear by this committee that led Facebook, Twitter,
and yes, Google to uncover the malicious activities of the
Russian-backed internet Research Agency activities on each of
your platforms.
Now each of you have come a long way with respect to
recognizing the threat. We've seen important action by your
companies to make political advertising more transparent--and
we discussed this yesterday--by complying with the terms
Senator Klobuchar and I put forward in the Honest Ads Act. In
addition, as the Chairman mentioned, since last September you
have identified and removed some bad actors from your
platforms.
The bad news, I'm afraid, is that there's still a lot of
work to do, and I'm skeptical that ultimately you'll be able to
truly address this challenge on your own. I believe Congress is
going to have to act.
First, on the disinformation front: Russia has not stopped.
Russian-linked information warfare exists today. Just recently,
we saw the two of you take action to take down suspected
Russian operations. We also know Microsoft uncovered Russian
attempts to hack political organizations and potentially
several political campaigns.
The Russians also continue to infiltrate and manipulate
American social media to hijack our national conversation.
Again, you've gotten better, and I'm pleased to see that you've
begun to take action, but also the Russians are getting better
as well. They have now become harder to track. Worse, now that
the Russian playbook is out there, other adversaries, as we saw
recently, like Iran, have joined the fray.
But foreign-based disinformation campaigns represent just a
fraction of the challenge before you. In the same way that
bots, trolls, fake pages, algorithmic gaming can be used to
spread fake news, these same tools can be used to assist
financial stock pumping fraud, to create filter bubbles and
alternative realities, to incite ethnic and racial violence,
and countless other misuses.
Imagine the challenge and damage to the markets if Ford's
communications from the Fed Chairman were leaked online. Or
consider the price of a Fortune 500 company's stock if a
dishonest short seller was able to spread false information
about the company's CEO or the effects of its products rapidly
online.
Russian disinformation has revealed a dark underbelly of
the entire online ecosystem, and this threatens to cheapen
American discourse, weaken privacy, erode truth, and undermine
our democracy on a previously unimagined scale. Worse, this is
only going to get harder as we move into artificial
intelligence, use of Deepfake technology.
During the 2016 election campaign, the Russians
demonstrated how bad actors can effectively marry offensive
cyber operations, including hacking, with information
operations. I'm afraid that we're on the cusp of a new
generation of exploitation, potentially harnessing hacked
personal information, to enable tailored and targeted
disinformation in social engineering efforts. That future
should concern us all.
As someone who was involved in the tech industry for more
than 20 years, I respect what this industry represents, and I
don't envy the significant technical and policy challenges you
face. But the size and reach of your platforms demand that we
as policy makers do our job to ensure proper oversight,
transparency, and protection for American users and our
democratic institutions.
The era of the Wild West in social media is coming to an
end. Where we go from here, though, is an open question. These
are complicated technological challenges, and Congress has at
times demonstrated that it still has some homework to do. I do
think this committee has done more to understand the threat to
our democracy posed by social media than any others, and I want
to commend my colleagues on this committee for tackling this
challenge in a bipartisan way.
As has been mentioned, this is our fourth public hearing on
the subject, and we've met behind closed doors countless times
with third-party researchers, with government officials, and
with each of the platforms. We've done the work, and we're
positioned to continue to lead in this space.
Again, as the Chairman has already indicated, today's
hearing is not about gotcha questions or scoring political
points. Our goal today is to begin to shape actual policy
solutions which will help us tackle this challenge.
Now, I've put forth some ideas that I'd like to get your
constructive thoughts on. For instance, don't your users have a
right to know when they're interacting with bots on your
platform? Isn't there a public interest in insuring more
anonymized data is available to help researchers and academics
identify the potential problems and misuse? Why are your terms
of service so difficult to find and nearly impossible to read,
much less understand? Why shouldn't we adopt ideas like data
portability, data minimization, or first-party consent? And
after witnessing numerous episodes of misuse, what further
accountability should there be with respect to the flawed
advertising model that you utilize?
Now these are just some of our ideas. We have received a
lot of positive feedback on some of these ideas from both
experts and users. We've also been accused of trying to bring
about the death of the internet. I'm anxious to hear your views
on our proposals and suggestions your teams can bring to the
table on this front.
We have to be able to find smart, thoughtful policy
solutions that get us somewhere beyond the status quo, without
applying ham-handed 20th-century solutions to 21st-century
problems. At the same time, we should be mindful to adopt
policies that do not simply entrench the existing dominant
platforms.
These are not just challenges for our politics or our
democracy. These threats can affect our economy, our financial
system, and other parts of our lives. I'm hopeful that we can
get there. I'm confident in American ingenuity. And I'm
optimistic that Congress led by this committee in a bipartisan
fashion can move this conversation forward.
I look forward to the discussion and appreciate the hearing
being called. Thank you, Mr. Chairman.
Chairman Burr. I thank the Vice Chairman. At this time, I'd
like to swear in our witnesses. If I could ask both of you to
stand and raise your right hand?
Do you solemnly swear to give this committee the truth, the
full truth and nothing but the truth so help you God?
[The witnesses answered in the affirmative.]
Please be seated. Ms. Sandberg, I'd like to recognize you
first and then Mr. Dorsey for any opening statement you'd like
to make. The floor is yours.
STATEMENT OF SHERYL SANDBERG, CHIEF OPERATING OFFICER, FACEBOOK
Ms. Sandberg. Thank you. Chairman Burr, Vice Chairman
Warner, and members of this select committee, thank you for
giving me the opportunity to speak with you today. My written
testimony goes into more detail about the actions we're taking
to prevent election interference on Facebook. But I wanted to
start by explaining how seriously we take these issues and talk
about some of the steps we're taking.
Free and fair elections are the foundation of any
democracy. As Americans, they are part of our national identity
and that's why it's incumbent upon all of us to do all we can
to protect our democratic process. That includes Facebook. At
its best, Facebook plays a positive role in our democracy,
enabling representatives to connect with their constituents,
reminding people to register and to vote, and giving people a
place to freely express their opinions about the issues that
matter to them.
However, we've also seen what can happen when our service
is abused. As a bipartisan report from this committee said,
Russia used social media as part of, and I quote: a
comprehensive and multi-faceted campaign to sow discord,
undermine democratic institutions and interfere in U.S.
elections and those of our allies.
We were too slow to spot this and too slow to act. That is
on us. This interference was completely unacceptable. It
violated the values of our company and of the country we love.
Actions taken show how determined we are to do everything we
can do to stop this from happening.
The threat we face is not new. America has always
confronted attacks from determined, well-funded opponents who
want to undermine our democracy. What is new is the tactics
they are using. To stay ahead, we all need to work together, as
Chairman Burr said: government, law enforcement, industry and
experts from civil society. And that is why I'm grateful for
the work this committee is doing.
At Facebook, we're investing in security for the long term.
As our defenses improve, bad actors learn and improve too, and
that's why security is never a finished job. We have more than
doubled the number of people we have working in safety and
security and we now have over 20,000 people and we are able to
view reports in 50 languages, 24 hours a day.
Better machine learning and artificial intelligence have
enabled us to be more proactive in finding abuse. In the first
three months of 2018 alone, over 85 percent of the violent
content we took down or added warning labels to was identified
by our technology before it was reported. These are expensive
investments, but that will not stop us because we know they are
critical.
Our first line of defense is finding and shutting down fake
accounts, the source of much of the inauthentic activity we see
on Facebook. Authenticity matters because people need to trust
that the content they're seeing is valid and they need to trust
the connections they make. We are now blocking millions of
attempts to register false accounts each and every day.
We're making progress on fake news. We're getting rid of
the economic incentives to create it and we're limiting the
distribution it gets on Facebook. We demote articles rated by
third-party fact-checkers as false. We warn people who have
shared them or who are about to share them, and we show them
related articles to give them more facts.
We've also taken strong steps to prevent abuse and increase
transparency in advertising. Today on Facebook, you can go to
any page and see all the ads that page is running, even if they
wouldn't be shown to you. For political and issue ads, you can
also see who paid for the ads, how much was spent, and the
demographics of the people who saw them.
We're also going to require people running large pages with
large audiences in the United States to go through an
authorization process and confirm their identity. These steps
won't stop everyone who's trying to game the system, but they
will make it a lot harder.
As these past few weeks and months have shown, this work is
starting to pay off. In July, we removed 32 pages and accounts
involved in coordinated, inauthentic behavior. In August, we
removed 650 pages and accounts that originated in Iran, as well
as additional pages and accounts from Russia. And just last
week, we took down 58 pages and accounts from Myanmar, many of
which were posing as news organizations.
We are focused, as I know you are, on the upcoming U.S.
midterms and on elections around the world. Our efforts in
recent elections from Germany, to Italy, to Mexico, to the
Alabama special Senate election, show us that the investments
we are making are yielding results. We also know, as Chairman
Burr said, that we cannot stop interference by ourselves. We're
working with outside experts, industry, partners and
governments, including law enforcement, to share information
about threats and prevent abuse.
We're getting better at finding and stopping our opponents,
from financially motivated troll farms to sophisticated
military intelligence operations. We don't have access to the
intelligence governments have access to, so we don't always
know exactly who is behind these attacks or their motives, and
that's why we will continue working closely with law
enforcement.
Chairman Burr, I want to thank you for your leadership.
Vice Chairman Warner, I want to thank you for your white paper,
which has so many ideas on how we can work together to
strengthen our defense. Senators, let me be clear, we are more
determined than our opponents and we will keep fighting.
When bad actors try to use our site, we will block them.
When content violates our policies, we will take it down. And
when our opponents use new techniques, we will share them so we
can strengthen our collective efforts.
Everyone here today knows that this is an arms race, and
that means we need to be ever more vigilant. As Chairman Burr
has noted, nothing less than the integrity of our democratic
institutions, processes, and ideals is at stake. We agree, and
we will work with all of you to meet this challenge.
Thank you.
[The prepared statement of Ms. Sandberg follows:]
[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]
Chairman Burr. Thank you, Ms. Sandberg. Mr. Dorsey, the
floor is yours.
STATEMENT OF JACK DORSEY, CHIEF EXECUTIVE OFFICER, TWITTER,
INC.
Mr. Dorsey. Thank you Chairman Burr, Vice Chairman Warner
and the committee for the opportunity--for the opportunity to
speak on behalf of Twitter to the American people. I look
forward to our conversation about the work we're doing to help
protect the integrity of U.S. elections and elections around
the world.
I am someone of very few words and typically pretty shy,
but I realize how important it is to speak up now. If it's OK
with all of you I'd like to read you something I personally
wrote as I considered these issues. I'm also going to tweet
this out now.
First, I want to step back and share our view of Twitter's
role in the world. We believe many people use Twitter as a
digital public square. They gather from all around the world to
see what's happening and have a conversation about what they
see. In any public space you will find inspired ideas and
you'll find lies and deception--people who want to help others
and unify, and people who want to hurt others and themselves,
and divide.
What separates a physical and digital public space is
greater accessibility and velocity. We're extremely proud of
helping to increase the accessibility and velocity of a simple,
free, and open exchange. We believe people would learn faster
by being exposed to a wide range of opinions and ideas, and it
helps make our Nation and the world feel a little bit smaller.
We aren't proud of how that free and open exchange has been
weaponized and used to distract and divide people and our
Nation. We found ourselves unprepared and ill-equipped for the
immensity of the problems that we have acknowledged: abuse,
harassment, troll armies, propaganda through bots and human
coordination, misinformation campaigns, and divisive filter
bubbles. That's not a healthy public square. Worse, a
relatively small number of bad faith actors were able to game
Twitter to have an outsized impact.
Our interests are aligned with the American people and this
committee. If we don't find scalable solutions to the problems
we're now seeing, we lose our business and we continue to
threaten the original privilege and liberty we were given to
create Twitter in the first place.
We weren't expecting any of this when we created Twitter
over 12 years ago. We acknowledge the real world negative
consequences of what happened and we take the full
responsibility to fix it. We can't do this alone and that's why
this conversation is important and why I am here.
We've made significant progress recently on tactical
solutions like identification of many forms of manipulation
intending to artificially amplify information, more
transparency around who buys ads and how they are targeted, and
challenging suspicious logins and account creation. We've seen
positive results from our work. We're now removing over 200
percent more accounts for violating our policies. We're
identifying and challenging 8 to 10 million suspicious accounts
every week, and we're thwarting over a half million accounts
from logging in to Twitter every single day.
We've learned from 2016, and more recently from other
nations' elections, how to protect the integrity of elections:
better tools, stronger policy, and new partnerships are already
in place. We intend to understand the efficacy of these
measures to continue to get better, but we all have to think a
lot bigger than decades past, today. We must ask the question,
what is Twitter incentivizing people to do, or not do, and why?
The answers will lead to tectonic shifts in Twitter and how our
industry operates. Required changes won't be fast or easy.
Today we're committing to the people and this committee to
do that work and do it openly. We're here to contribute to a
healthy public square, not compete to have the only one. We
know that's the only way our business thrives and helps us all
defend against these new threats.
In closing, when I think of my work, I think of my mom and
dad in St. Louis, a Democrat and a Republican. For them,
Twitter has always been a source of joy, a source of learning,
and a source of connection to something bigger than themselves.
They're proud of me, proud of Twitter, and proud of what made
it all possible. What made it possible was the fact that I was
born into a Nation built by the people for the benefit of the
people--where I could work hard to make something happen which
was bigger than me. I treasure that and will do everything in
my power to protect it from harm. Thank you.
[The prepared statement of Mr. Dorsey follows:]
[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]
Chairman Burr. Jack, thank you very much for that testimony
and I might add that the Vice Chairman and I commented as you
grow older, you will find a need for a bigger device to go to
your notes on than that small one. We have a hard time with the
small devices.
For members, we will do seven minute question rounds today.
For planning purposes, we will break at approximately 10:45 for
five minutes just to let our witnesses stretch and take a
breath. And we will limit today's hearing to one round. We'll
try to accommodate any members that might be caught in the
Judiciary Committee but want to try to get back, but I know
that they've got their own challenges. With that, I would
recognize myself for seven minutes.
This question is to both of you. How would you define
social media for this committee and more importantly for the
American people? And I will start with you, Ms. Sandberg.
Ms. Sandberg. Social media enables you to share what you
want to share when you want to share it, without asking
permission from anyone. And that's how we meet our mission,
which is giving people a voice. And I think what's more
important than just the content people share, is the
connections they make. Social media enables people to celebrate
their birthdays. In the last year, people have raised $300
million on Facebook on birthday funders for nonprofits they
care about. Safety check: Millions of people in the worst
circumstances of their lives have let their loved ones know
they're safe. And small businesses to grow. All around the
country I meet with small businesses, from a woman making
dresses in her living room and selling them on Instagram, to a
local plumber, who are able to find their customers on Facebook
and then able to grow and hire people and live their American
dream.
Chairman Burr. Jack.
Mr. Dorsey. I believe it's really important to--to
understand how the people see it. And we believe that the
people use Twitter as they would a public square and they often
have the same expectations that they would have of any public
space. For our part, we see our platform as hosting and serving
conversations. Those conversations are in the public. We think
there's a lot of benefit to those conversations being in the
public, but there's obviously a lot of risks as well.
We see that news and entertainment are actually byproducts
of public conversation. And we see our role as helping to not
only serve that public conversation so that everyone can
benefit, even if they don't have a Twitter account, but also to
increase the health of that conversation as well. And in order
to do that, we need to be able to measure it. We need to
understand what healthy participation looks like in a public
square, and we need to amplify that. And more importantly, we
need to question a lot of the fundamentals that we started with
12 years ago in the form of incentives. When people use our
product every single day--when they open our app up--what are
we incentivizing them to do? Not telling them what to do, but
what are we actually incentivizing them to do? And that
certainly speaks to the buttons that we have in our service,
all the way to our business model.
Chairman Burr. Ms. Sandberg, this question is for you. One
root problem that we see is that users don't truly understand
the types of data that are being collected on and off your
platform. How is that data shared with advertisers or others to
deliver targeted advertising and what vetting, if any, do you
do on targeted advertising to prevent hostile actors from
targeting your users for their products?
Ms. Sandberg. Senator, it's a really important question
because it goes to the heart of our service. We sell ads and we
use information that people share with us or share with third-
party sites to make those ads relevant to them. But privacy and
advertising are not at odds. In fact, they go together. When
people share information with us, we do not give it to
advertisers without their permission. We never sell data. And
they have control over the information we use.
Chairman Burr. Again for both of you, and I'll start with
you, Mr. Dorsey. What's your company's ability to collaborate
with other social media companies in this space?
Mr. Dorsey. We have a real openness to this and we have
established a more regular cadence with our industry peers. We
do believe that we have an opportunity to not only create more
transparency with an eye towards more accountability, but also
a more open way of working and a way of working that, for
instance, allows for a review period by the public on how we
think about our policies.
But more so, taking some of the lessons that we have
learned and benefited from in the open-source software space to
actually think about developing our policies, our enforcement,
and also our products going forward. We've been experimenting a
little bit with this recently, but we would like to be a
company that is not only hosting an open conversation but is
also participating in that open conversation. So, we're more
than open to more collaboration, and not just with our industry
peers but with scholars, academics, and also our government
partners.
Chairman Burr. Thank you.
Ms. Sandberg.
Ms. Sandberg. I think our collaboration has greatly
increased. We've always worked closely with law enforcement and
we continue to do that and particularly the FBI's new task
force. We've always shared information with other companies but
I think we are doing better and we can continue to do better.
Mr. Chairman, you noted in your opening remarks that some
of the tips we got came from a private security firm. In our
mind that's the system working. Our opponents are very well-
funded. They are very organized, and we are going to get those
tips from law enforcement, from each other, from private firms.
And the faster we can collaborate, the faster we share those
tips with each other, the stronger our collective defenses will
be.
Chairman Burr. Last question from the Chair--again for both
of you and I'll go in reverse--you first, Ms. Sandberg. If a
foreign-influence campaign is detected among your platforms, is
there a defined process by which other platforms are alerted to
the campaign that you've discovered?
Ms. Sandberg. Our security teams have been in close contact
and so right now when we find something, we are reaching out to
our companies--other companies to do it and working more
closely together.
We've been talking about how, I think, there's still room
for improvement there. I think we can do more to formalize the
process. We've had a series of meetings and I think we're going
to continue to work and we can do better.
Chairman Burr. Mr. Dorsey.
Mr. Dorsey. This is not something we want to compete on. We
hosted our peer companies at our offices just in the past two
weeks on this very topic and helping to increase our cadence of
meeting and also what we can share. If there were an
occurrence, we would immediately look to alert our peer
companies and this committee and our government law enforcement
partners.
Chairman Burr. Thank you for that. Let me just say in
closing that I hope both of you, if you see impediments that
exist in your ability to notify or to collaborate as it relates
to nefarious actors, that you'll certainly make this committee
aware in cases where we can help. With that, Vice Chairman.
Vice Chairman Warner. Thank you, Mr. Chairman. As I
indicated in my opening statement, I hope we can move forward
on the policy discussion, so I'd like to get your thoughts on
some of the ideas I and others have suggested, and I want to
start with you, Mr. Dorsey.
I think after some initial false starts, it does really
appear that you have committed to a shift in your company's
culture with respect to the safety and security on your
platform. Obviously, I have been impressed by some of the
increasing efforts you've taken. A question I have, though, is
that obviously on your platform there are a lot of automated
accounts or bots, and there's nothing inherently good or bad
about an automated account. As a matter of fact, there are
certain very good things that come out of some of these
automated accounts. But, do you believe that an individual
Twitter user should have the right to know when he or she is
being contacted, whether that contact is initiated by a human
being or a bot?
Mr. Dorsey. I do believe that first and foremost, anyone
using Twitter has the right to more context around not only the
accounts that they're seeing, but also the information.
Vice Chairman Warner. Would that go as far as actually
having a policy on your platform indicating--I wouldn't ask you
to take them down--but at least allowing the user to know
whether that contact was initiated by a human being versus a
machine?
Mr. Dorsey. As far as we can detect them. We can certainly
label and add context to accounts that come through our API.
Where it becomes a lot trickier is where automation is actually
scripting our website to look like a human actor. So as far as
we can label--and we can identify these automations--we can
label them, and I think that is useful context and it's an idea
that we have been considering over the past few months. It's
really a question of the implementation, but we are interested
in it and we are going to do something along those lines.
Vice Chairman Warner. It's not going to solve the problem,
but I do think giving that indication to users would allow them
then perhaps to make a little more judgment. Because we had,
for example, back in early August, we had a panel of experts,
and they were saying that some of the content--in terms of
political content, I'm not talking about total tweets--but
total political content was 25 to 30 to 1 on the far left and
far right generated by either foreign actors or automated
accounts. And my question is: Doesn't that volume on the
extremes drown out real conversation and political conversation
amongst Americans, regardless of where they fall on the
political spectrum?
Mr. Dorsey. It does, in the shared areas of Twitter. So
there are two main categories of usage in Twitter. One, is the
people you follow, and those Tweets end up in your timeline.
Two, are the more common shared spaces, like Search, Trends,
and also Replies. That's where anyone could interject
themselves, and that's where we see the most gaming of our
systems, and that's where we've also made the most progress in
terms of identifying these patterns and shutting them down
before they spread too far. That is independent of our work on
automation, because we're seeing the same patterns through
human coordination as well.
Vice Chairman Warner. I appreciate your comments about the
willingness to notify a user whether it's a human being or a
machine contacting you. I also think that there's room for
improvement on some of the high volume Twitter accounts, to
really do a little bit of extra examination.
Ms. Sandberg, let me move to you. Obviously, in a digital
economy, I think data increasingly represents the single
greatest asset you have. Obviously it's a part of the
advertising model that you've created.
But I think most users are actually pretty much in the dark
about how much data is actually being collected on them, what
it's actually worth. I think as we've seen from other fields,
like health care, the fact that we have such a lack of price
transparency really makes health care reform really
challenging.
I think some of that lack of price transparency and value
within social media also exists, so I'd like to first of all
ask, does a Facebook user have a right to know what information
you are collecting about that user?
Ms. Sandberg. Yes, and we really agree with you that people
who use Facebook should understand what information is being
used, how it's used, and the controls they have. We've worked
hard to simplify this. We've put out things like privacy
shortcuts, which show you all your settings in one place, and
something called download your information, where you can
download all of your information in a portable way and be able
to take it with you and see what it is.
Vice Chairman Warner. I understand, and I think you're
making progress there, but again, if a user has that
information, he or she may not know the value. Wouldn't it be
actually helpful to your user to actually be able to then put
some valuation on the data you're collecting from the user and
publish that in a way so that people actually know what their
information is worth?
Ms. Sandberg. Mr. Vice Chairman, I think this is one of the
proposals you laid out in your white paper, and like all of
this, you know, we don't think it's a question of whether
regulation--we think it's a question of the right regulation
that supports users, is transparent, and doesn't squash
innovation. And we're happy to work with you on the proposal.
Vice Chairman Warner. Well, I just think it's that more
price transparency is always better, and I think this would be
something that would help users sort through. There was another
question that we've talked in the past about: Is there
anything, even with a willing user, are there any rights or
details about an individual user that they should not be able
to give up or consent to having used?
Ms. Sandberg. I'm sorry, I don't understand the question.
Vice Chairman Warner. My question is this: At some point,
are there certain pieces of personalized information that a
user shouldn't be able to voluntarily give to an enterprise
like yours or Twitter?
Ms. Sandberg. I think there are, and I think there are many
ways users have control over what they do. I also think there
are probably corner cases of law enforcement holds or security
matters where information is critically important.
Vice Chairman Warner. I just wonder whether--just a
question of whether you can consent away all of your rights--
ought to be something we ought to have a discussion on. I've
only got a few more seconds.
Let me ask, Ms. Sandberg, you made mention in your opening
testimony the fact that sometimes political actors are using
the platforms really to incent violence. I think you made at
least some mention of Myanmar, where we've obviously seen a
great tragedy take place there, where hundreds of thousands of
Rohingya Muslims are fleeing in many ways. The U.N. High
Commissioner has said that fake accounts on Facebook have
incented that violence.
Do you believe that Facebook has both a moral obligation
and potentially even a legal obligation to take down accounts
that are actually incentivizing violence?
Ms. Sandberg. I strongly believe that. In the case of
what's happened in Myanmar, it's devastating, and we're taking
aggressive steps and we know we need to do more. Probably the
most important thing we've done is ramped up our ability to
review reports in Burmese.
Vice Chairman Warner. I appreciate your comment that
Facebook would have both a moral and legal obligation, so
sorting through what that would look like so that if there were
other platforms that weren't being as responsible, there ought
to be some sanctions. So I look forward to working with you on
that issue as well.
Thank you, Mr. Chairman.
Chairman Burr. Senator Risch.
Senator Risch. Thank you. Thank you both for being here
today. This is, I think, the third hearing we've held over the
last year or so--fourth--the Chairman says the fourth--that
we've had on this issue.
I think the problem is really well laid out. We've spent
hours and hours and hours talking about this and what the
issues are and what the problems--I'm still not hearing what--
very specifically how we're getting after this. I know there're
some things being done. I tend to agree with you that no matter
what's done, as long as these platforms are there, there's
going to be people finding their way into it to do bad things.
And obviously, everybody wants to get that reduced as much as
possible.
And I'm glad to hear that you and the entire industry are
trying to do something about this. The entity up here that I
serve in, there are lots of people that would love to help you
run your organizations through what we call the regulatory
process. That isn't all of them, obviously, and hopefully it
isn't even a majority of them, but there will be--and you've
already seen efforts in that regard--but you're going to have
to do things yourselves to try to get around this so that we
don't have the horrible things happen that spawn that type of
regulation.
I want to drill down a little bit. In each of your
companies, who sets these standards or the description of what
a coordinated manipulation or inauthentic behavior is? What
entity do you have in each of your companies who make these
determinations?
Ms. Sandberg, let me start with you.
Ms. Sandberg. Our policy team is setting those, and our
security team is finding them. And coordinated inauthentic
behavior means behavior on our site that's inauthentic, so
people are not representing themselves to be who they are to
be. And coordinated means they are coordinating it, and they
can be coordinating with authentic actors and coordinating with
inauthentic actors. Both are unacceptable.
Senator Risch. When the team is sitting there meeting, is
there generally a unanimity amongst them on something--a fact
situation comes in front of them. Is this something that is
easy to recognize--people are unanimous about it--or do you
wind up with debates as to whether or not a certain platform
should be shut down?
Ms. Sandberg. I think on a lot of issues we face like hate
speech, there's broad debate. When it comes to what is an
inauthentic actor, which is a fake account posing as someone,
they're hard to find. But once we find them, we know what they
are.
Senator Risch. And what about--the Chairman referred to
standards in his opening statement. Who sets these standards,
the same committee?
Ms. Sandberg. The same group of people.
Senator Risch. And are they published, so that a user can
look at that? Well, give me some examples of standards that are
unacceptable.
Ms. Sandberg. In the coordinated inauthentic behavior or in
general?
Senator Risch. In general.
Ms. Sandberg. Yes, so we publish our community standards
comprehensively. And what that does is define what's permitted
on Facebook and what's not permitted on Facebook. So some
examples are, bullying is not permitted, hate is not permitted,
language that leads to violence is not permitted, and this is
published in detail publicly.
Senator Risch. Mr. Dorsey, where's your company on these
things?
Mr. Dorsey. So, we have a team called Trust and Safety who
is responsible for designing and writing these policies that
reports up to our lead of legal and safety, and--and our
compliance teams which report directly to me.
Senator Risch. I'd like to ask both of you: One of the
things this committee wrestles with frequently when it comes to
privacy issues and those kinds of things is the difference
between a U.S. citizen and a non-U.S. citizen. And under U.S.
law, they can be treated differently under different
circumstances.
Do your companies make any distinction between a U.S.
citizen versus a non-U.S. citizen? And I guess, now I'm more
focusing in on the kind of behavior we saw where elections are
attempted to be manipulated and--and that sort of thing. Ms.
Sandberg, let's start with you. Does your company make a
distinction as they're weighing the activity of certain actors?
Ms. Sandberg. So for political and issue ads, we are now
going through a verification process. And in order to run those
in the United States, people have to verify that they are
legally able to do that. So that's one area where we would
distinguish.
Senator Risch. And what does that mean, legally able to do
that? If a citizen of another country, any other country,
decides they want to say something about a U.S. election, are
they disqualified from doing that with your company?
Ms. Sandberg. In the free content--so what their posts are
to their friends and family or publicly--people are allowed to
talk about any issues in any country, as long as they're not
crossing over into the areas we discussed that aren't allowed,
like hate and bullying. In advertising, in U.S. elections, you
have to be a U.S. citizen.
Senator Risch. Mr. Dorsey.
Mr. Dorsey. We have very similar policies and we do segment
them by advertising and also the more organic social creation
of content as well.
We don't always have an understanding of where an account
is located. We have to infer this oftentimes. And this is where
we do get a lot of help from our law enforcement partners. It
is not only to understand where some of these threats are
coming from, but also the intent. And the faster that we get
that information, the faster that we can act.
Senator Risch. One of the concerns that I have--and I
appreciate that explanation--but what we've seen on this
committee, and have actually seen in other contexts, is that in
today's world it is so easy to either employ or even
impersonate a U.S. citizen to do something in a given context.
Do you have difficulties in that regard?
Ms. Sandberg. Well, finding inauthentic behavior is a
challenge and I think you're seeing us put real resources to
bear. This is why we're investing so heavily in people and
technology. This is why we're investing in programs like
verification.
I think the other step we're taking here is around
transparency. So being able to see if people bought political
ads, where they're located, being able to see who's running a
page; these are steps we think are really important for helping
us find what--to your point--can be very difficult things to
find.
Senator Risch. Mr. Dorsey, briefly.
Mr. Dorsey. We've decided to focus a lot more on the
behavioral patterns that we're seeing across the network. While
we can't always recognize in real-time where someone might be
coming from or if they were--if they are representing someone
who does not exist, we can see common patterns of behavior and
utilizing the network to spread their information.
So we have been building a lot of our machine learning and
deep learning technology to recognize these patterns and shut
them down before they spread too quickly. And then, also, link
them to other accounts that demonstrate similar patterns. And
we've gotten a lot more leverage out of that in terms of
scalability than working on systems to identify whether it's a
fake profile or not.
Senator Risch. Interesting, thank you.
Chairman Burr. Senator Wyden.
Senator Wyden. Thank you, Mr. Chairman.
Mr. Chairman, I want to thank you and Senator Warner for
your kind comments about John McCain. And what is not often
remembered is John McCain wrote some of the really important
rules of the road for the internet when he was Chairman of the
Commerce Committee. And it was always bipartisan, so I very
much appreciate both of you mentioning our wonderful friend,
John McCain.
And Ms. Sandberg, Mr. Dorsey, welcome and I've enjoyed
visiting with you. Let me go right to the question that is
foremost on my mind, and that is consumer privacy as a national
security issue.
Technology companies like yours hold vast amounts of very
private information about millions of Americans. The prospect
of that data being shared with shady businesses, hackers, and
foreign governments is a massive privacy and national security
concern. Russians keep looking for more sophisticated ways of
attacking our democracy.
Personal data reveals not just your personal and political
leanings, but what you buy, even who you date. My view is
personal data is now the weapon of choice for political
influence campaigns. And we must not make it easier for our
adversaries to seize these weapons and use them against us.
So I'd like to see if we could do a yes or no on this. And
I wrote it because I think we can. My view is, from this point
on, beefing up protections and controls on personal privacy
must be a national security priority. I'd like a yes or no, Ms.
Sandberg.
Ms. Sandberg. Yes.
Senator Wyden. Mr. Dorsey.
Mr. Dorsey. Yes.
Senator Wyden. Okay. Let me turn now to a question based on
a lot of analysis my office has done and you all have talked to
us about. We have reviewed Facebook privacy audits required by
the 2011 consent agreement after your company was found to use
unfair and deceptive practices.
One section of the audits deals with how Facebook shared
the personal information of Americans with smart phone
manufacturers. These included the Chinese companies Huawei and
ZTE. I found portions of this audit very troubling and the
findings could affect many Americans. I believe, Ms. Sandberg,
the American people deserve to see this information. Will you
commit this morning to making public the portion of your audits
that relate to Facebook's partnerships with smart phone
manufacturers?
Ms. Sandberg. Senator, I really appreciate the question and
the chance to clarify this issue because it's really important.
With regards to the audits, our third-party auditor, PWC, does
audits on a rolling basis every two years, but they're
continual. They are given to us. We have shared them with the
FTC voluntarily and we will continue to do that.
I can't commit right in this moment to making that public
because a lot of that has sensitive information which could
help people game the system, but we will certainly work with
you to see what disclosures would be prudent. But----
Senator Wyden. Let's do this. Because that's a constructive
answer and I've got other things I've got to cover. I'm just
going to assume you will work with this. We understand the
question of redaction on sensitive national security matters.
Can you get back to me within a week with respect to how
Facebook will handle what I think is troubling information?
Ms. Sandberg. We're going to get back to you as quickly as
possible. We can definitely prioritize this request. So we'll
do it as fast as we can depending on the volume of requests
everyone has.
Senator Wyden. Thank you. And look, so you all know where
I'm going with this. To me, protecting data privacy has to be a
higher tier issue in terms of national security. It's going to
be the foundation of the legislation that I've talked to both
of you about. So that's why I feel strongly and I think your
answer is constructive and I hope we can get that quickly.
What I also want to get to with you, Ms. Sandberg, is the
issue of micro targeting to discourage voting. This is one of
the most powerful tools in the propaganda arsenal. Going after
individual Americans with ads and really lasering in on the
ability to affect political campaigns. It's certainly been used
in the past with the Russians to discourage minority Americans
from voting. Would Facebook's current policies prohibit using
micro targeting to discourage voting?
Ms. Sandberg. Senator, we feel very strongly about this.
There is a long history in this country of trying to suppress
civil rights and voting rights and that activity has no place
on Facebook. Discriminatory advertising has no place on
Facebook.
Senator Wyden. So what are you doing to prohibit this
micro-targeting? I mean what about ads that share false
information about the date of the election or the location of a
polling place or ads that tell people they can vote with a text
message from their phone. You have said that it's unacceptable
to target minorities and others, but I really need to drill
down more deeply in knowing, because I think this is a
primary--we can get bipartisan agreement on. What do you do to
deal with micro targeting?
Ms. Sandberg. So with everything when we're looking for
abuse of our systems and things that are against our policies,
we have a combination of people reviewing ads, and we have a
combination of automated systems and machine learning that help
us find things and take them down quickly.
Senator Wyden. OK, I'll hold the record open for that.
Could I have, say within a week, a written answer that would
get into some of those specifics?
Ms. Sandberg. We're going to get you answers to your
questions as quickly and thoroughly as we can.
Senator Wyden. Good. My last question deals with foreign
governments aiding hoaxes and misinformation and I'd like to
get both of you, in fact. Why don't you start with this Mr.
Dorsey?
Do either of you or your companies have any indication that
Iran, Russia, or their agents have supported, coordinated with,
or attempted to amplify the reach of hoaxes?
Mr. Dorsey.
Mr. Dorsey. Of hoaxes?
Senator Wyden. Yes.
Mr. Dorsey. We certainly have evidence to show that they
have utilized our systems and gamed our systems to amplify
information. I'm not sure in terms the definition of hoax in
this case, but it is likely.
Senator Wyden. Okay.
Ms. Sandberg.
Ms. Sandberg. Just two weeks ago, we took down 650 pages
and accounts from Iran. Some were tied to state-owned media and
some of them were pretending to be free press, but they weren't
free press. So it depends on how you define a hoax, but I think
we're certainly seeing them use misinformation to campaign----
Senator Wyden. My time is up. The only other area I'm going
to want to explore with you is, we've got to deal with this
back and forth between the private sector and the government.
Very often, we ask you all about things you're doing and you
say we need the government to also help us get to A, B, C, and
then the government says the same thing about you. We'll want
to explore that. Thank you Mr. Chairman for the extra time.
Chairman Burr. Senator Rubio.
Senator Rubio. I want to thank you both for being here.
First of all, there's an empty chair next to you from
Google. They're not here today and maybe it's because they're
arrogant or maybe it's because there's a report that as of last
night--this was posted at 3:36 yesterday--this group went on
basically pretending to be Kremlin-linked trolls. They did
everything. They used the details of the Internet Research
Agency, which is a Kremlin-linked troll farm, and were able to
buy ads online and place them on sites like CNN, CBS This
Morning, HuffPost, The Daily Beast, so I'm sure they don't want
to be here to answer these questions.
But I thank you both for being here. I was happy to read in
your opening statement, Ms. Sandberg, that you talk about our
democracy, our democratic process. You acknowledge
responsibility for protecting our process. And you talked about
our adversaries, clearly linking the company to the values and
the importance of this country and I think in acknowledgment
that your company would not exist were it not in the United
States, because of the freedoms that we have.
Twitter didn't go as far, but you did describe yourself as
a global town square--but you did say that you want to support
free and open democratic debate. You did refer to our democracy
and you did say that Twitter was built on the core tenet of
freedom of expression, which is a very important core tenant.
Here is why this is relevant, because we're here today
because we learned--and we've learned the hard way--that social
media was largely seen as a tool for incredible good. Also,
what makes it good can be manipulated by bad actors to do harm.
And that's what happened. We have all learned that the hard
way.
And so what we're asking you to do, and I think what you've
agreed to do, is to use the powers that you have within your
platforms to crack down on certain users who are hostile
actors, who are using disinformation or misinformation or hate
speech for the purposes of sowing discord, or interfering in
our internal affairs--and that's a positive.
Here's the problem though: we have to start thinking about
what happens when an authoritarian regime asks you to do that
because their definition of disinformation or misinformation
could actually be the truth. Their discord, or what they define
as discord, would be things like defending human rights.
Interfering in their internal affairs, they would define as
advocating for democracy. And the reason why I think that
answering that question is so important is because it's going
to define what your companies are. Are your companies really
built on these core values, or are they global companies, like
all these other companies that come around here, who see their
number one obligation to make money and therefore market access
irrespective of the price they have to pay to do so?
So, for example, in 2016 the New York Times reported that
Facebook was working on a program to restrict stories from
showing up in newsfeeds based on the user's geography. The
story implies--and I know that it hasn't been implemented--but
it implies that that was being used in order to potentially try
to get back into China, but any authoritarian government could
try to use that tool.
Vietnam, by the way, where you do operate, has a new law
beginning on 2019 January 1st that will require you to store
user data inside the country and hand over that data, to the
government, of users suspected of anti-state activity,
including spreading news that may impede Hanoi or hurt the
economy, for example, democracy activists.
Twitter has a policy of accommodating countries that have
different ideas about the contours of freedom of expression by
selectively blocking tweets and accounts. For example, one of
the countries you complied with is Pakistan, who has asked you
to block sites for blasphemy. The blasphemy--647 cases of
blasphemy over a ten-year period from 1986 to 2007. Fifty
percent of those cases were on non-Muslim Pakistanis--in a
country three percent non-Muslim.
One high-profile case is Asia Bibi, who has been sentenced
to death after a personal dispute over drinking water with a
group of women. They accused her of insulting the prophet.
She's arrested, imprisoned, sentenced to death. Not relevant to
Twitter but relevant to the blasphemy laws that Pakistan has
asked you to comply with.
Turkey has requested that you block over 12,000 accounts.
Since 2014, you've blocked over 700. Many of them are
journalists. One of them is an NBA player, Enes Kanter. Russia
blocked almost 80 accounts as of last check. You complied with
that. One of them was a pro-Ukrainian account in 2014.
And so here's why all of this is relevant. I guess the
first question for Facebook is: These principles of our
democracy--do you support them only in the United States or are
these principles that you feel obligated to support around the
world?
Ms. Sandberg. We support these principles around the world.
You mentioned Vietnam. We do not have servers in Vietnam. And
with very minor exceptions of imminent threats that were
happening, we've never turned over information to the
Vietnamese government, including political information.
Senator Rubio. And you never will?
Ms. Sandberg. We would not.
Senator Rubio. You would not agree to do so in order to
operate?
Ms. Sandberg. We would only operate in a country when we
can do so in keeping with our values.
Senator Rubio. And that would apply to China as well?
Ms. Sandberg. That would apply to China as well.
Senator Rubio. Thank you. And on Twitter, how is blocking
the account of journalists or an NBA player in keeping with the
core tenant of freedom of expression?
Mr. Dorsey. We enacted a policy some time ago to allow for
per-country content takedown. Meaning that within the
boundaries of that nation, the content would not be able to be
seen but the rest of the world can see it. And that's important
because the world can still have a conversation around what's
happening in a market like Turkey. And also, we have evidence
to show that a lot of citizens within Turkey access that
content through proxies and whatnot, as well.
So, we do believe--and we have fought the government--the
Turkish government--consistently around their requests and
oftentimes won. Not in every case, but oftentimes have made
some moves. So we would like to fight for every single person
being able to speak freely and to see everything, but we have
to realize that it's going to take some bridges to get there.
Senator Rubio. Well, because a Twitter spokesman in
response to a Buzzfeed article--I think about two years ago--
here's the quote defending this policy. It said, ``Many
countries including the United States have laws that may apply
to tweets and/or Twitter account content.'' And then you went
on to say what you said, ``On our continuing efforts to make
services available to users everywhere et cetera.'' You would
agree that there's no moral equivalency between what we're
asking you to do here and what Turkey has asked you to do, or
other countries have asked you to do, in that same realm?
Mr. Dorsey. We do have to comply with the laws that govern
us within each one of these nations, but our ideals are similar
and our desires----
Senator Rubio. Whose ideals are similar? I'm sorry.
Mr. Dorsey. The company's.
Senator Rubio. Are similar to who?
Mr. Dorsey. Similar to how we were founded and where we
were founded in this country.
Senator Rubio. I guess my point is, you're not arguing
though that what we're asking you to do here--on this
misinformation against foreign efforts to interfere in our
elections--is the same as what Turkey or other authoritarian
regimes have asked you to do abroad, against political
opponents of theirs. They're not morally equivalent, these two
things?
Mr. Dorsey. Correct.
Senator Rubio. Thank you.
Chairman Burr. The Chair will recognize Senator Heinrich
for questions and then members should know that we will take a
short recess, no more than five minutes, and then reconvene.
Senator Heinrich.
Senator Heinrich. Thank you, Mr. Chair, and thank you both
for being here. I think we've learned quite a bit over the
course of the last couple of years. I think it would be an
understatement to say that we were all caught flat-footed in
2016: social media platforms, the intelligence community, this
committee, government as a whole.
Obviously, we want to learn from that and what I'd like to
start with is to ask from each of you, since 2016 your
platforms have been used throughout the course of a number of
subsequent elections--elections in France, in Germany, and
other Western allies across Europe.
What have you learned from those consequential elections
after 2016 and how has that informed your current posture in
terms of how you're gaining transparency into this activity? Go
ahead.
Ms. Sandberg. Senator, I think we've learned a lot and I
think we're going to have to continue to learn because as we
learn, our opponents learn, and we have to keep up. We're
working on technology and investments in people making sure
fake news is disseminated less on the platforms--transparency
actions and taking down bad actors.
And we've seen everywhere, from Mexico to Brazil to other
places around the world, these same techniques deployed
differently and each time we see it, I think we get smarter. I
think we see the new threat and I think we're able to connect
the dots and prevent those threats going forward.
Senator Heinrich. Mr. Dorsey.
Mr. Dorsey. We've also learned a lot from elections around
the world, most recently the Mexican election. We have opened a
new portal to cover that election, that allows any journalist
or government law enforcement to actually report any suspicious
behavior very quickly to us, so we can take more actions.
Otherwise, we have been investing in artificial
intelligence and machine learning models to, again, recognize
the patterns of behavior because we believe this is where the
greatest leverage will come from, recognizing how people
artificially amplify information and shutting it down before it
spreads into the shared spaces of Twitter and more broadly into
someone's replies to a tweet.
Senator Heinrich. I want to get to the basic issue of
whether our incentives in this case are aligned to deal with
these challenges. If your users were to lose confidence in your
platforms, in the authenticity of what you, Mr. Dorsey, called
a public square--I might call it a digital public square--I
assumed there would be very serious economic implications for
your companies. Do you think the--the incentives have aligned
for platform providers of all types in the digital space, to
want to get at these issues, and have a plan, and be able to
respond in real time?
Ms. Sandberg and then you, Mr. Dorsey.
Ms. Sandberg. Absolutely. Trust is the cornerstone of our
business. People have to trust that what they see on Facebook
is authentic. People have to trust that this is a positive
force for democracy and the things they care about. And so this
has been a huge issue for us and that's why we're here today
and that's why we're going to keep working to get ahead of
these threats and make sure we can minimize all of this
activity.
Mr. Dorsey. Our incentives are aligned but I do believe it
goes a lot deeper than just the alignment of our company
incentives with this committee and the American people. I
believe we need to question the fundamental incentives that are
in our product today.
Every time someone opens up our service, every time someone
opens up our app, we are implicitly incentivizing them to do
something or not to do something. And that extends all the way
to our business and those answers that we get from asking that
question are going to create massive shifts in how Twitter
operates and I also believe how our industry operates. So what
worked 12 years ago does not work today--it hasn't evolved fast
enough--but I think it is a layer--many, many, many, many
layers deeper than the surface symptoms that we often find
ourselves discussing.
Senator Heinrich. Ms. Sandberg, you mentioned a number of
things that would violate your standards, for example, hate
speech, advocacy of violence. What about when you were dealing
with real people, authentic users, intentionally spreading
false information? And obviously there are huge free speech
implications there. But, for example, what if a real person, a
U.S. citizen, says that victims of the mass shootings were
actually actors? Would that violate your standards and if the
answer is no, how should we, and by we, I mean government and
industry, deal with those very real challenges?
Ms. Sandberg. Well let me start by saying I find claims
like that personally, unbelievably upsetting. If you've been a
victim or a parent of a victim, they deserve all our full
support. And finding a line between what is hate speech and
what is misinformation is very, very difficult, especially if
you're dedicated to expressing free expression, and sometimes
free expression is expressing things you strongly disagree
with.
In the case of misinformation, what we do is we refer it to
third-party fact-checkers. We don't think we should be the
arbiter of what's true and what's false, and we think that's
really important. Third-party fact-checkers then mark it as
false. If it's marked as false, we dramatically decrease the
distribution on our site. We warn you if you're about to share
it. We warn you if you have shared it and, importantly, we show
related articles next to that so people can see alternative
facts.
The fundamental view is that bad speech can often be
countered by good speech, and if someone says something is not
true and they say it incorrectly, someone else has the
opportunity to say, actually you're wrong. This is true and
that's what we're working on through our systems.
Senator Heinrich. I think one of the things we found in
2016 is that we didn't have the transparency and the literacy
to do what you just pointed out there: to counter false speech
with accurate speech to understand how this speech was
propagating in the digital public space.
What more do you think we should be doing to simply make
the public more literate about the fact that this information
warfare is very real? It's going on all the time. It's not fake
news. It's not a hoax. It's something we're all going to have
to deal with, that our kids, even playing platforms like
Pokemon Go, may have to--have to deal with as well.
Do either of you have a quick opinion on that? And then my
time will be expired. I apologize, Mr. Chair.
Mr. Dorsey. I believe we need to point to where we see
healthy participation and clearly mark what is healthy and what
is unhealthy. And also realize that not everyone is going to
choose healthy participation in the short term. But how do we
encourage healthy participation in order to increase the reach
and also increase the value of what they're giving to that
digital public square.
Chairman Burr. This hearing stands in a recess subject to
the call of the Chair.
[Whereupon the hearing recessed at 10:51 a.m. and
reconvened at 11:01 a.m.]
Chairman Burr. I'd like to call the hearing back to order.
The chair would recognize Senator Collins for questions.
Senator Collins. Thank you, Mr. Chairman. First let me
thank you both for being here and also to express my outrage
that your counterpart at Google is not at the table as well.
Mr. Dorsey, as of January of this year, Twitter has taken
down more than 3,800 Russian IRA accounts that by Twitter's own
estimate reached approximately 1.4 million people. One of those
accounts purported to be under the control of the Tennessee
GOP, although it was not. It was a Russian IRA account. It had
more than 140,000 followers and would sometimes spread
conspiracy theories and false claims of voter fraud.
My question to you is: Once you have taken down accounts
that are linked to Russia, these impostor accounts, what do you
do to notify the followers of those accounts that they have
been following or engaged in accounts that originated in
Russia, and are not what they appear to be?
Mr. Dorsey. Thank you for the question. We simply haven't
done enough. So in this particular case, we didn't have enough
communication going out in terms of what was seen and what was
tweeted, and what people are falling into.
We do believe transparency is a big part of where we need
the most work and improvement, and it's not just with our
external communications, it's actually within the product and
the service itself.
We need to meet people where they are, and if we determine
that people are subject to any falsehoods or any manipulation
of any sort, we do need to provide them the full context of
that. And this is an area of improvement for us and something
that we're going to be diligent to fix.
Senator Collins. I think this is critically important. If a
follower just gets a message that says this Twitter account is
no longer available, that does not alert the individual that he
or she has been receiving messages--tweets--from a Russian
entity whose goal is to undermine public confidence in elected
officials and our democratic institutions.
So I really think we need something more than even the
tombstone, or something else. We need to tell people that they
were taken in or victims--innocent victims--of a foreign
influence campaign.
Ms. Sandberg, let me ask you this same question. What is
Facebook doing?
Ms. Sandberg. We agree with you that people need to know,
so we've been discussing these publicly, as well as in specific
cases notifying people. So we notified people directly if they
had liked--or had liked the original IRA accounts.
Most recently when there was an event that was going to be
happening in Washington that inauthentic accounts--we notified
all the people who either RSVP'd to that event, or who said
they were interested in possibly going to that event.
Senator Collins. Thank you. That was the Night to Defeat
the Right, or something like that, as I recall.
Mr. Dorsey, back to you. Clemson University researchers and
others have shown that these Russian IRA accounts target
specific leaders and social movements across the political
spectrum. And again, the goal of the Russians, the Iranians--
anyone else who is involved in this influence campaign--is to
undermine the public's confidence in political leaders and
weaken our democratic institutions and turn us against one
another.
Well, I learned not from Twitter but from Clemson
University that I was one of those targeted leaders and that
there were 279 Russian-generated tweets that targeted me that
had gone to as many as 363,000 followers. So why doesn't
Twitter notify individuals like me that we have been targeted
by foreign adversaries? I shouldn't find out from looking at
Clemson University's database and working with their
researchers. It seems to me that once you determine that, you
should notify the people who are the targets.
Mr. Dorsey. I agree. It's unacceptable. And as I said
earlier, we want to find ways to work more openly, not just
with our peer companies but with researchers and universities
and also law enforcement because they all bring a different
perspective to our work, and can see our work in a very
different light. And we are going to do--we're going to do our
best to make sure that we catch everything and we inform people
when it affects them. But, we are not going to catch
everything. So it is useful to have an external partnership and
work with them to make sure that we're delivering a message in
a uniform manner where people actually are, without requiring
them to find a new channel to get that information.
This is where a lot of our thinking is going and a lot of
our work is going. But we recognize we need to communicate more
directly where people are on our service, and we also recognize
that we're not going to be able to catch everything alone, so
we need to develop better partnerships in order to do that.
Senator Collins. I would close my questioning by
encouraging both of you to work more closely with academia,
with our government. The Clemson University researchers have
done extraordinary work, but they have said that they've been
provided data that is only within the last three years, which
does not allow them to do the kind of analysis that they'd like
to do and that's probably because of the new European Union
privacy laws. But the EU has provided research exemptions. So I
hope that you will commit to providing data that goes beyond
that three year window to researchers who are looking into
Russian influence efforts on your platforms. Thank you.
Chairman Burr. Senator Harris.
Senator Harris. Thank you, Mr. Chairman, for accommodating
me. I'm in another hearing as you know. Good morning, and to
the invisible witness, good morning to you. So I have a few
questions for Ms. Sandberg. On November 2, 2017, your company's
general counsel testified in front of this Intelligence
Committee on Russian interference, and I asked a few questions.
I asked how much money did you make, and this is of the
representative from both Facebook and Twitter--both of your
general counsels were here. And I asked how much money did you
make from legitimate advertising that ran alongside the Russian
propaganda. The Twitter general counsel said, quote, ``We
haven't done the analysis but we'll follow-up with you and work
on that.'' And the Facebook general counsel said the same is
true for Facebook.
Again, I asked Facebook CEO Mark Zuckerberg on April 10,
2018, and he said that, quote, ``Internet Research Agency, the
Russian firm, ran about $100,000 worth of ads.'' Following the
hearing, I asked Facebook the same question in writing, and on
June 8, 2018, we received a response that said, quote, ``We
believe the annual revenue that is attributable to inauthentic
or false accounts is immaterial.''
So my question is: What did you mean by immaterial? Because
I'm a bit confused about the use of that term in this context.
Ms. Sandberg. Thank you for the question.
Again we believe the total of the ad spending that we have
found is about $100,000. And so the question you're asking is
with the inorganic content, I believe, what is the possible
revenue we could have made? So here's the best way I can think
of to estimate that, which is that we believe between 2015 and
2017, up to 150 million people may have seen the IRA ads or
organic content in our service. And the way our service works
is, ads don't run attached to any specific piece of content,
but they're scattered throughout the content. This is
equivalent to .004 percent of content in news feed and that was
why they would say it was immaterial to our earnings.
But I really want to say that from our point of view,
Senator Harris, any amount is too much.
Senator Harris. If I may, just so I'm clear about your
response--so are you saying that then the revenue generated was
.004 percent of your annual revenue? Of course that would not
be immaterial.
Ms. Sandberg. Again, the ads are not attached to any piece
of content so----
Senator Harris. So what metric then? Just help me with
that. What metric are you using to calculate the revenue that
was generated, associated with those ads? And what is the
dollar amount that is associated then with that metric?
Ms. Sandberg. The reason we can't answer the question to
your satisfaction is that ads are not--organic content--ads
don't run with inorganic content on our service, so there is
actually no way to firmly ascertain how much ads are attached
to how much organic content. It's not how it works.
In trying to answer what percentage of the organic----
Senator Harris. But what percentage of the content on
Facebook is inorganic?
Ms. Sandberg. I don't have that specific answer, but we can
come back to you with that.
Senator Harris. Would you say it's the majority?
Ms. Sandberg. No. No.
Senator Harris. An insignificant amount? What percentage?
You must know.
Ms. Sandberg. If you ask about our inauthentic accounts on
Facebook, we believe at any point in time it's 3 percent to 4
percent of accounts, but that's not the same answer as
inorganic content because some accounts generate more content
than others.
Senator Harris. I agree. So what percentage of your content
is inorganic?
Ms. Sandberg. Again, we don't know. I can follow up with
the answer to that.
Senator Harris. Okay, please. That would be great. And then
your company's business model is obviously--it's complex but
benefits from increased user engagement and that results of
course in increased revenue. So, simply put, the more people
that use your platform, the more they are exposed to third-
party ads, the more revenue you generate. Would you agree with
that?
Ms. Sandberg. Can you repeat? I just want to make sure I
got it exactly right.
Senator Harris. So the more user engagement will result--
and the more then that they are exposed to third-party ads--the
more that will increase your revenue. So the more users that
are on your platform----
Ms. Sandberg. Yes. Yes. But only I think when they see
really authentic content. Because I think in the short run and
over the long run it doesn't benefit us to have anything
inauthentic on our platform.
Senator Harris. That makes sense. In fact, the first
quarter of 2018, the number of daily active users on Facebook
rose 13 percent, I'm told. And corresponding ad revenue grew by
half to $11.79 billion. Does that sound correct to you?
Ms. Sandberg. Sounds correct.
Senator Harris. And then would you agree that--I think it's
an obvious point--that the more people that engage on the
platform, the more potential there is for revenue generation
for Facebook?
Ms. Sandberg. Yes, Senator. But again, only when the
content is authentic.
Senator Harris. I appreciate that point. And so a concern
that many have is how you can reconcile an incentive to create
and increase your user engagement when the content that
generates a lot of engagement is often inflammatory and
hateful.
So, for example, Lisa-Maria Neudert, a researcher at Oxford
and Internet Institute, says, quote, ``The content that is the
most misleading or conspiratorial, that's what's generating the
most discussion and the most engagement, and that's what the
algorithm is designed to respond to.''
My concern is that according to Facebook's community
standards, you do not allow hate speech on Facebook. However,
contrary to what we've seen, on June 28, 2017, a ProPublica
report found that Facebook's training materials instructed
reviewers to delete hate speech targeting white men but not
against black children because black children are not a
protected class. Do you know anything about that, and can you
talk to me about that?
Ms. Sandberg. I do. And what that was, I think, a bad
policy that's been changed, but it wasn't saying that black
children--it was saying that children--it was saying that
different groups weren't looked at the same way, and we've
fixed it.
Senator Harris. But isn't that the concern with hate,
period? That not everyone is looked at the same way?
Ms. Sandberg. Well, hate speech is against our policies and
we take strong measures to take it down. We also publish
publicly what our hate speech standards are. We care
tremendously about civil rights. We have worked very closely
with civil rights groups to find hate speech on our platform
and take it down.
Senator Harris. So when did you address that policy? I'm
glad to hear you have. When was that addressed?
Ms. Sandberg. When it came out--and again, that policy was
a badly written, bad example, and not a real policy.
Senator Harris. The report that I'm aware of was from June
of 2017. Was the policy changed after that report or before
that report from ProPublica?
Ms. Sandberg. I can get back to you on the specifics of
when that would have happened.
Senator Harris. You're not aware of when it happened?
Ms. Sandberg. I don't remember the exact date.
Senator Harris. Do you remember the year?
Ms. Sandberg. Well, you just said it was 2017.
Senator Harris. So do you believe it was 2017 that the
policy changed?
Ms. Sandberg. It sounds like it was.
Senator Harris. Okay. And what is Facebook's official
stance on then hate speech regarding so-called, and legally
defined, unprotected classes, such as children?
Ms. Sandberg. Hate speech is not allowed on our platform
and hate speech is, you know, important in every way. And we
care a lot that our platform is a safe community. When people
come to Facebook to share, they're coming because they want to
connect on the issues that matter to them.
Senator Harris. So, have you removed the requirement that
you will only protect with your hate speech policy those
classes of people that have been designated as protected
classes in a legal context? Is that no longer the policy of
Facebook?
Ms. Sandberg. I know that our hate speech policies go
beyond the legal classifications and they are all public and we
can get back to you on any of that. It's all publicly
available.
Senator Harris. Thank you so much. Thank you, Mr. Chairman.
Chairman Burr. Senator Blunt.
Senator Blunt. Thank you, Chairman. Mr. Dorsey, Wired
magazine last week had an article that said you'd admitted
having to rethink fundamental aspects of Twitter. Would that be
an accurate reflection of where you've been the last year?
Mr. Dorsey. Yes. We are rethinking the incentives that our
service is giving to people.
Senator Blunt. And what would be the biggest area where
you're trying to rethink how you thought this was going to work
out and the way it's turned out to be?
Mr. Dorsey. Well--and this is pretty far-reaching--so we're
still in the process of doing this work, but when we created
the service 12 years ago, we had this concept of followers. And
we made the number of followers big and bold and a very simple
but noticeable font.
And just that decision alone has incentivized people to
want to grow that number, to increase that number. And the
question we're now asking is, ``Is that necessarily the right
incentive? Is the number of followers you have really a proxy
for how much you contribute to Twitter and to this digital
public square?'' And we don't believe it is. But that's just
one question. The way we lay out our buttons on the bottom of
every tweet in a reply and a retweet and a like, that also
implies an incentive and a point of view that we're taking that
we want to encourage people to do.
So as we think about serving the public conversation, as we
think about our singular priority of increasing the health of
that public conversation, we are not going to be able to do
long-term work unless we are looking at the incentives that our
product is telling people to do every single day.
Senator Blunt. All right, that's helpful. Thank you.
Senator Collins asked her last question--I didn't really quite
get the answer to that question. But I think what she was
asking is a question I had also, which was: In the interest of
transparency and public education and looking at things
available to researchers and policy makers, are you willing to
archive suspended accounts so that people can look back at
those? And would that be a period of, I think, three years was
part of the question she asked. Give me a little better, more
specific answer. You didn't have time to answer that, and I'd
like you to have time to answer that.
Mr. Dorsey. We are looking at things like a transparency
report. We put out a transparency report around terrorism, but
we're looking at expanding that transparency report around
suspensions of any account.
We are still coming up with the details of what this will
look like and what it will include.
Senator Blunt. As opposed to just a transparency report,
are you willing to archive some of this where you may not be
reporting on it at the time, but someone could look three years
down the road and try to do an analysis of why that information
was out there the way it was and how it fit into your overall
policy of taking whatever action you're taking?
Mr. Dorsey. I think it's a great idea to show the
historical public record. We just need to understand what the
legal implications are, and we can get back to you on that.
Senator Blunt. Yes, I may come back with a question if I
have time on legal implications, generally. I think for both of
your companies, who have been pretty forward-leaning in the
last couple of months as this conversation has moved pretty
dramatically, the business implications, the liability
implications of what we're asking you to do are pretty great.
Well, let me see if I can get a couple of Facebook
questions in first. Ms. Sandberg, does Facebook differentiate
between foreign and domestic influence operations when deciding
whether to take down a page or remove an account from the
platform?
Ms. Sandberg. Our focus is on inauthenticity, so if
something is inauthentic, whether it's trying to influence
domestically or trying to influence on a foreign basis--and
actually a lot more of the activity is domestic--we take it
down.
Senator Blunt. You take it down indiscriminately, whether
it's a foreign influence or--or a domestic influence?
Ms. Sandberg. And you saw that with the IRA. With the IRA
accounts, the original ones for our election were targeted at
the United States, but then there were another 270 accounts
that were almost all targeted in Russia or at Russia--for
Russian speakers and nearby languages. So a lot of those were
domestic, and those are down.
Senator Blunt. Well, it's been mentioned several times, and
I think appropriately so, Google is not here today. But the two
of you are, and Ms. Sandberg, again, just what seems like a
long time ago, but only a few months, since Mr. Zuckerberg was
here testifying before Congress. It seems like to me that
Facebook has been pretty active in finding and taking down
things that should not have been out there: the recent Iranian
takedown, the Russian things that have been taken down.
Do you want to talk a little about what's the big challenge
about being at the forefront of trying to figure this out from
a business perspective or a liability perspective, either one?
Then I'm going to come to Mr. Dorsey with the same question.
Ms. Sandberg. Well I really appreciate what you said,
because we have been investing very heavily in people, in our
systems, in decreasing the dissemination of fake news, in
transparency, and I think that's what you're seeing pay off.
I think we've all said, in the private meetings we had as
well as this public discussion, that tighter coordination
really helps us. If you look at our recent takedowns, some of
it was information we found ourselves, some of it were hints we
got from law enforcement, some of it is information we can
share with other companies.
And so this is a big threat, and our opponents are going to
keep getting better and we have to get better. We have to stay
ahead. And the more we can all work together the better off
we're going to be, and that's why I really appreciate the
spirit with which this hearing this morning is taking place.
Senator Blunt. And how does the takedown, the practice
work, where legitimate accounts are sold then maybe--and
repurposed by others? What are you looking at there as a
challenge?
Ms. Sandberg. So our policy is inauthenticity. If you are
an inauthentic account, if you are pretending to be someone
you're not, you come down. If you have touched the account of
someone who is authentic, then we would leave the authentic
account up, but in cases like I was answering with Senator
Collins, if you are an authentic person who RSVP'd to an event
that's not authentic, we would let you know.
Senator Blunt. Okay, thank you for that. Okay, Mr. Dorsey,
back to that other question. From a business and legal
liability standpoint, what's the downside of being out there
where you are now trying to every day implement policies that
nobody's ever implemented before?
Mr. Dorsey. I think there are a number of short-term risks,
but you know, we believe that the only way that we will grow
and thrive as a company is by increasing the health of this
digital public square that we're helping to build. We also
benefit, as Sheryl mentioned, from tighter collaboration and
tighter partnership. We've really strengthened our partnership
with our government agencies since 2016.
There are a few areas that we would like to see more
strength. We would like a more regular cadence of meetings with
our law enforcement partnerships. We would love to understand
the secular trends that they are aware of, and seeing in our
pure companies or other mediums, or more broadly that would
inform us about how to act much faster. And we would appreciate
as much as we can consolidating to a single point of contact,
so that we are not bouncing between multiple agencies to do our
work.
So that is what we've found in attempting to do a lot of
this new policy and work, in terms of partnership, but
ultimately it comes back to: we need to build our technologies
to recognize new patterns of behavior and new patterns of
attack, and to understand what they actually mean, and then
ideally get some help from our law enforcement partners to
understand the intent and to understand the motivations behind
it.
Senator Blunt. Thank you, Mr. Dorsey. I'm sure my time is
up. Thank you, Chairman.
Chairman Burr. Senator King.
Senator King. Thank you, Mr. Chairman, and I want to also
thank our witnesses. And thank you to your companies and your
policy makers for making really great strides in the last year.
As many of the people have talked about, we were all on our
heels a year ago on this subject. And this has emerged as one
of the most important parts of this committee's investigation.
I try to focus on what we're after here. And we're after
the heart of democracy. Ms. Sandberg, you said the heart of
democracy was free and fair elections. I would argue that the
heart of free and fair elections is information. And that's
really what we're talking about: getting information to people
in a democratic setting. And also on all kinds of other topics,
birthdays and everything else, but that's what we're talking
about here.
There are three ways to defend ourselves it seems to me.
One is better consumer discrimination about what they're
seeing. The second is deterrence, which hasn't been mentioned
here, that our adversaries need to understand that there's a
price to be paid for trying to manipulate our society and our
democracy. And the third is technical, and that's mostly what
we've been talking about.
I had an experience, ironically, a couple of months before
the 2016 election, meeting here in this building with a group
of people from Lithuania, Estonia, and Latvia, who have been
experiencing Russian interference with their elections and
their propaganda, their information for years. And I said,
``How do you defend yourself?'' You can't unplug the internet.
You can't turn off the TV station. The most interesting thing
they said was, universally, the best defense is for the people
to know it's happening.
And I would like from each of you some thoughts and
hopefully a commitment to educating your users about the
potential for abuse of the very medium that they're putting
their trust in.
Ms. Sandberg.
Ms. Sandberg. We really agree with you. And we've done this
broadly and we're going to continue to do more. So we've worked
on media literacy programs. We've worked on programs in public
service announcements around the world that help people
discern--this is real news, this is not--and help people be
educated. I think one of the most important things we're doing
is that once a piece of content has been rated as false by our
third-party fact-checkers--if you're about to share it, we warn
you right there. Hey this has been rated as false. And so, you
are educated as you are about to take that critical step.
Senator King. And Mr. Dorsey, I hope you're doing the same
to educate your users as to the potential that they can be
misled on your platform.
Mr. Dorsey. Yes. And to be frank, we haven't done a good
job at this in the past. And I think the reason why is because
we haven't met our customers where they are, in terms of
actually when they're using the product and adding more context
there.
We do benefit on Twitter that we have this amazing
constituency of journalists globally using our service every
single day, and they often, with a high degree of velocity,
call out nonfactual information. We don't do a great job at
giving them the best tools and context to do that work. And we
think there's a lot of improvements we can make to amplify
their content and their messaging so that people can see what
is happening with that content.
Senator King. If that can be amplified and underlined, it
can become a self-healing process, whereby the response
immediately responds to false or misleading information.
Deterrence, I'm not going to spend a lot of time on, except
to say that many of us believe that one of the great gaps in
our defenses against election interference and interference in
our democracy is the fact that our adversaries feel no pain if
they do so--that we have to develop a doctrine of cyber
deterrence just as we have doctrines of military deterrence.
And that's a gap, and that's something that we're working on
both here and at Armed Services, other places.
Let me talk about the technical for a minute. How about
feedback from users? And Ms. Sandberg, you testify that you
have third-party fact-checkers. Also, would it be useful to
have more in the way of ratings? And, you know, the eBay
sellers--you have rating process and number of stars, and those
kinds of things. Is there more you could do there to alert
people as to the validity and the trustworthiness of what
they're seeing?
Ms. Sandberg. Senator, the most important determinant of
what anyone sees on Facebook are decisions they make. So I
choose my friends, you choose yours. I choose the news
publications I follow, you choose yours. And that's why your
news feed is so different from mine. And so, yes, if you don't
want to follow someone, if you don't want to like a page, we
encourage you to do that. We also make it very easy to unfollow
on our site. So if I don't believe what you're saying anymore,
I don't have to receive your----
Senator King. But I'm talking about alerting a viewer or a
reader to something that's come across on their newsfeed that
has been found manifestly false or misleading: a banner, a
note, a star.
Ms. Sandberg. We do that through related articles. We note
this has been rated as false, and here's a related article
which would give you other facts that you could consider.
Senator King. One of the things that we've been talking
about here, and Senator Rubio has been a leader in discussing
this, is what we call Deepfake, as I'm sure you're aware, the
ability to manipulate video to the point where it basically
conveys a reality that isn't real.
Is there a technological way that you can determine that a
video has been manipulated in that way and tag it? So that
people on Facebook, if they see a video that it'll be tagged:
warning, this has been manipulated in a way that may be
misleading. That's a question you may want to take under
advisement. But it seems to me, again, this is an area--this is
a new area that's going to get more and more serious, I'm
afraid. And again, what I'm trying to do is give the consumer
the maximum amount of information.
Ms. Sandberg. We agree with you, Deepfakes is a new area
and we know people are going to continue to find new ones. And
as always, we're going to do a combination of investing in
technology and investing in people so that people can see
authentic information on our service.
Senator King. As you're thinking about these cures, I hope
you'll continuously come back to the idea that what we need to
do is give people more information. I must say, I'm a little
uncomfortable with where the line is between taking down
misleading or fake information and taking down what someone
else may consider legitimate information in the marketplace of
ideas. Jefferson said we can tolerate error, as long as truth
is left free to combat it. We have to be sure that we're not
censoring. But at the same time, we're providing our customers,
our users--your users with information that they can--the
context, I think, is the word you use--they can have context
for what it is that they're seeing.
I'd hate to see your platforms become political in the
sense that you're censoring one side or the other of any given
debate.
Mr. Dorsey.
Mr. Dorsey. So yes, we absolutely agree. As we are building
a digital public square, we do believe expectations follow
that. And that is a default to freedom of expression and
opinion. And we need to understand when that default interferes
with other fundamental human rights such as physical security
or privacy. And what the adverse impact on those fundamental
human rights are.
And I do believe that context does matter in this case. We
had a case of voter suppression around 2016 that was tweeted
out. And we are happy to say that organically, the number of
impressions that were calling it out as fake were eight times
that of the reach of the original tweet. That's not to say that
we can rely on that, but asking the question how we make that
more possible, and how we do it at velocity is the right one to
ask.
Senator King. That's the self-healing aspect. Thank you
both very much. And if you have further thoughts as you're
flying home, about technical ways you can increase the
information available to your users through tags, ratings,
stars, whatever, please share them with us and we'll look
forward to working with you on this problem that is one that's
important to our country. Thank you very much.
Chairman Burr. Senator Lankford.
Senator Lankford. Thank you, Mr. Chairman. I want to follow
up on a statement that Senator King was mentioning as well
about Deepfakes. That's something I've spoken to both of you
about before in the past. It is a challenge for us and I would
just reiterate some of the things that he was saying publicly.
When it's the possibility and now the opportunity to be able to
create video that looks strikingly real, but none of it is
actually real--all of it is computer-generated--that is a very
different day for video-sharing in the days ahead. And I know
as you all have attacked issues like child pornography and
other things on your platforms in the past, you all will
aggressively go after these things. We're just telling you
we're counting on it because Americans typically can trust what
they see, and suddenly in video they can no longer trust what
they see because the opportunity to be able to create video
that's entirely different than anything in reality has now
actually come. And so I appreciate your engagement on that.
And I want to talk to you a little bit, Mr. Dorsey, about
following up some of the things that Senator Blunt had
mentioned as well about suspended accounts. When you suspend an
account, obviously there's information that's still there. Do
you archive all of that information to be able to maintain for
a suspended account that this is an account that we determine
is either from a foreign actor or hostile actor or is
inappropriate--not an authorized user? Is that something you
hold that information, so you can maintain it?
Mr. Dorsey. I need to follow up with you on the exact
details of our policies, but I believe we do, especially in
regards to any law enforcement action.
Senator Lankford. Terrific. For Facebook, what is the
practice when you suspend an account and say this is not an
authorized user or we think this is a foreign or hostile user?
Ms. Sandberg. If we have any suspicion that it's a foreign
or hostile user, we would keep the information to be able to do
further investigation.
Senator Lankford. So then the question is, is the
investigation internal for you all? Or obviously if law
enforcement subpoenas that and comes to you and says I have a
subpoena to come get that information, that's a whole different
issue. But is that something you do in your own investigation?
Because as I'm sure you've seen in the past, some users will
create a fake account or some sort of hostile account. That
comes down, they'll create another one, and then there's some
similarities in where they go and directions and relationships.
Do you maintain that data to be able to make sure that
you're well prepared and educated for when they may come back
to be aware of that again? For Twitter what is that, Mr.
Dorsey?
Mr. Dorsey. So we do, do our own internal investigations
and we are benefited every time our peers recognize something,
and we do share that data so that we can check our own systems
for similar vectors or similar accounts. And also work with law
enforcement to understand the intent. If there is a request to
allow an account to lay dormant by law enforcement, we will
allow that to happen and work with them to make sure that we
are tracking it accordingly.
Senator Lankford. Mr. Dorsey, the main thing I'm trying to
identify though is, let's say it happened in 2017. You identify
an account that you suspended and said this is your problem
area or an unauthorized user, whatever it may be.
You take that account off, do you maintain that
information? And so a year later if somebody comes back on with
a similar profile you can still track it and say, this is the
same as what we've seen before and it's going to take
additional steps for you to get back on board or ways to be
able to track their initial connections?
Mr. Dorsey. I'm sorry, yes. We do maintain that information
and we have a ban evasion policy. So if someone is trying to
evade a ban or suspension, no matter what the timeframe, we can
take action on those accounts as well.
Senator Lankford. Okay.
Ms. Sandberg.
Ms. Sandberg. If we have any suspicion that this would be
engaged in foreign or domestic inauthentic activity or we have
law enforcement interaction on it, we would keep that
information.
Senator Lankford. Okay. Mr. Dorsey, you and I have spoken
on this as well about data and the business model for both of
you is obviously--it's a free platform for everyone to use--but
obviously data and advertising and all those things are very
helpful just in keeping your business open and keeping your
employees paid. That's a given, and everyone understands that
when they join that platform and that conversation. But for
data in particular, how do you make sure that anyone who
purchases into data or gets access to that uses it for its
stated purpose, rather than using it to either sell to a third
party or to open up as a shell company, and say they're using
it for one purpose but they're actually using it for a foreign
purpose or direction to be able to track real-time activity of
Americans? How do you assure that companies that are purchasing
into that opportunity to have that data are actually fulfilling
and using it as they stated they would?
Mr. Dorsey. Well, there's a few things here. First and
foremost, we're a little bit different than our peers and that
all of our data is public by default. So when we sell data,
what we're selling is speed and comprehensiveness. So you're
actually purchasing either insights or a real-time streaming
product. In order to purchase that you have to go through a
very strict know-your-customer policy that we enact and then we
audit every single year. If we have any indication that there
is suspicious activity happening, that is an opportunity for us
to reach out to law enforcement with the sole purpose of trying
to understand the intent. That is the thing that we are not
always going to be able to infer from us looking at the
relationship.
You mentioned setting up companies that potentially are in
front of governments. That is not information that we would
necessarily have and that is where we are dependent upon the
intelligence to inform us so that we can take stronger action.
Senator Lankford. So, how do you determine or what
relation--is it an initial relationship but there's not a
follow up after that rapid access as you dictate on that? After
that is determined, is there any way to check in on those
companies to be able to make sure they're actually fulfilling
their terms of service?
Mr. Dorsey. Absolutely. And we do it every year on a
regular basis. But if we see anything suspicious at any point
in time, we'll reach out directly.
Senator Lankford. Ms. Sandberg, tell me a little bit about
WhatsApp? WhatsApp has been a feature of Facebook for a while.
How is the encryption going on that? What's the relationship
now with WhatsApp and what do you anticipate in the days ahead?
Ms. Sandberg. We are strong believers in encryption.
Encryption helps keep people safe. It secures our banking
system, it secures the security of private messages, and
consumers rely on it and depend on it. And so we're very
committed to encryption in WhatsApp and continuing to protect
the data and information of our users.
Senator Lankford. So that encryption is end-to-end at this
point still on the WhatsApp platform?
Ms. Sandberg. We'll get back to you on any technical
details, but to my knowledge, it is.
Senator Lankford. Thank you. I yield back.
Chairman Burr. Senator Manchin.
Senator Manchin. Thank you, Mr. Chairman. And Ms. Sandberg
and Mr. Dorsey, I want to thank both of you for being here. And
I grew up in an age without computers and social media so I'm
trying to get acclimated the best I can. I have seen how
they've been used by my children and grandchildren and how much
it helps connect people. I see an awful lot of good.
I also have concerns with internet and social media have
been--how it's been used against us. And I think you're hearing
concerns from all of my fellow colleagues up here. It's an
attempt to divide Americans, change our way of life, change our
democracy as we know it, and it can be very devastating.
In my little State of West Virginia--my beautiful little
State of West Virginia, with all the wonderful people--has been
hit extremely hard by illicit drugs and pharmaceutical opiates.
According to the recent Wired article, Eileen Carey spent three
years regularly reporting accounts illegally selling opiates on
Instagram. And the practice was widespread on Facebook and
Twitter, as well.
In many ways, the tools used by opiate dealers are similar
to those adopted by other bad actors including Russia, target
the vulnerable with ads that are easily circumventing the
platforms, filters, and oversights, and using hashtags to gain
attention of those interested. Last November, Facebook CEO Mark
Zuckerberg said learning of the depths of the crisis was the
biggest surprise and really saddening to see. But it still took
months to take measures to correct the problem while other
people were still dying.
According to the U.S. Code 230, formerly known as a
Communications Decency Act of 1996, online service providers
shall not be held civically liable for content that a third
party posts on their platform, and they shall not be treated as
a publisher or speaker of the content.
If we look at the example of drug overdose deaths, many
prosecutors are increasingly treating the deaths as a homicide
and looking to hold someone criminally accountable. There are
now laws devised to hold drug dealers responsible for the death
of victims using drugs they provided and, in some cases, they
are charging friends, partners, siblings of the deceased.
So my question to both of you would be: I've heard of a
report that details the way drug dealers continue to use your
platforms for illegal drug sales. To what extent do you bear
responsibility for the death of a drug user if they overdosed
on drugs received through your platform?
Either one. I know it's a tough one.
Ms. Sandberg. I'm happy to go.
Senator Manchin. Yes.
Ms. Sandberg. This is really important to us. The opioid
crisis has been devastating, and takes the lives of people in
our country and around the world. It's firmly against our
policies to buy or sell any pharmaceuticals on Facebook, and
that includes the opioid drugs. We rely on a combination of
machines and people reporting to take things down, and I think
we've seen marked improvements.
We also took an additional step recently which is very
important which is, we're requiring treatment centers who want
to buy ads to be certified by a respected third party because
another one of the problems has been that some treatment
centers are actually doing harm, and so we're requiring
certification before they can purchase ads and they can try to
reach people for treatment.
Mr. Dorsey. This is also prohibited on our service and we
do have a responsibility to fix it anytime we see it. And we
are looking deeply at how this information spreads, and how the
activity spreads so that we can shut it down before it spreads
too far.
Senator Manchin. I know I asked a tough question. It was,
do you all feel any responsibility because there has been a lot
of people that have been affected, and a lot of people have
died receiving information on how to obtain drugs through your
all's platform?
So I would go another step further, just like we passed
FOSTA and SESTA--FOSTA was the Fight Online Sex Trafficking
Act, and stop enabling--and SESTA was the Stop Enabling Sex
Traffickers Act. We passed bills that held you liable and
responsible. Don't you think we should do the same with opiate
drugs and the way they're being used in your platform? Would
you all support us doing that?
Mr. Dorsey. We're certainly open to dialogue around CDA and
the evolutions of it. We benefit from a lot of the protections
it gives in order for us in the first place to take actions on
the content within our service. The only reason we're able to
even speculate that we can increase more health in a public
square is because of CDA 230. So we need to finely balance what
those changes are and what that means.
Senator Manchin. Well, did it change your all's approach of
how you use your platforms with the changing of Code 230?
Mr. Dorsey. We have to do that independent of changes to
230.
Ms. Sandberg. These things are against our policies, and we
want them off and we want to take all measures to get them off.
The Safe Harbor of 230 has been very important in enabling
companies like ours to do proactive enforcement, look for
things proactively, without increasing our liability. And so,
we'd want to work very closely on how this would be enacted.
Senator Manchin. Final question to both of you. Why are you
not doing business in China?
Mr. Dorsey. We are blocked in China.
Ms. Sandberg. We are as well.
Senator Manchin. You're blocked? For what reasons?
Ms. Sandberg. The Chinese government has chosen not to
allow our service in China. I think it happened on the same
day.
Senator Manchin. Did you all not accept, basically, the
terms of how you do business in China? Or you're just blocked
from coming in to it? Or did you not agree? Did they give you a
chance, or--? I'm saying other social platforms seem to be
adapting and going in there.
I know a lot of our drugs--a lot of the fentanyl and all
that--is coming from China, and we're trying to shut that down.
But it was interesting to me that you all both have been
blocked. And I would assume you didn't agree to their terms?
Mr. Dorsey. I don't know if there's any one particular
decision point around understanding what the terms might be in
our particular case. But when we were blocked, we decided that
it wasn't a fight worth fighting right now, and we have other
priorities.
Senator Manchin. Are you still looking to do business
there?
Ms. Sandberg. There was no particular time. You know, we've
been open about the fact that our mission is to connect the
world. And that means, it's hard to do that without connecting
the world's largest population. But in order to go into China,
we would have to be able to do so in keeping with our values.
And that's not possible right now.
Senator Manchin. Thank you.
Thank you, Mr. Chairman.
Chairman Burr. Senator Cotton.
Senator Cotton. I want to commend both of you for your
appearance here today, for what was no doubt going to be some
uncomfortable questions. And I want to commend your companies
for making you available. I wish I could say the same about
Google.
I think both of you, and your companies, should wear it as
a badge of honor that the Chinese Communist Party has blocked
you from operating in their country. Perhaps Google didn't send
a senior executive today because they've recently taken actions
such as terminating cooperation that they had with the American
military on programs like artificial intelligence that are
designed not just to protect our troops and help them fight and
win our country's wars, but to protect civilians as well. This
is at the very same time that they continue to cooperate with
the Chinese Communist Party on matters like artificial
intelligence, or partner with Huawei and other Chinese telecom
companies that are effectively arms of the Chinese Communist
Party. And credible reports suggest that they are working to
develop a new search engine that would satisfy the Chinese
Communist Party's censorship standards after having disclaimed
any intent to do so eight years ago.
Perhaps they didn't send a witness to answer these
questions because there is no answer to those questions, and
the silence we would hear right now from the Google chair would
be reminiscent of a silence that that witness would provide.
So I just want to ask both of you, would your companies
ever consider taking these kinds of actions that privilege a
hostile foreign power over the United States and especially our
men and women in uniform.
Ms. Sandberg.
Ms. Sandberg. I'm not familiar with the specifics of this
at all, but based on how you're asking the question, I don't
believe so.
Mr. Dorsey. Also no.
Senator Cotton. So thank you for that answer. Mr. Dorsey,
let's turn to Dataminr, which is one of the services that
provides basically all of Twitter's data. The last time we had
an executive from Twitter before this committee in an open
setting, I asked about reports that Dataminr had recently
ceased its cooperation with the Central Intelligence Agency, at
the same time it continued to cooperate with Russia Today and
other proxies of Russian intelligence services.
I have since seen reports that Dataminr no longer
cooperates with Russia Today or any other proxy of Russian
intelligence services. Is that correct?
Mr. Dorsey. That is correct.
Senator Cotton. Did you make that decision personally?
Mr. Dorsey. No, we have a long-standing term against
utilizing public Twitter data for ongoing 24/7 surveillance.
Senator Cotton. And that's why you've decided to cease
cooperation with the Russian government or proxies like Russia
Today?
Mr. Dorsey. No. That's a different matter. This is in
regards----
Senator Cotton. Could you explain why you ceased that
cooperation then, or that relationship with, Russia Today and
other Russian intelligence proxies?
Mr. Dorsey. When we learned of the link of Russia Today and
Sputnik, we ceased to allow them to be an advertiser on the
platform. We calculated the amount of advertising they did on
the platform is $1.9 million and we donated that to civil
liberties nonprofits.
Senator Cotton. Would you now reconsider the decision to
cease your cooperation with the Central Intelligence Agency or
other American intelligence agencies?
Mr. Dorsey. We are always open to any legal process that an
agency would present us, so we don't believe it necessary. This
is a global policy around surveillance in general and real-time
surveillance. I will state that all this information, because
Twitter is public by default, is available to everyone by just
going to our service.
Senator Cotton. You see a difference between cooperating
with the United States government and the Russian government or
the Chinese government?
Mr. Dorsey. Do I see a difference? I'm not sure what you
mean.
Senator Cotton. Is Twitter an American company?
Mr. Dorsey. We are an American company.
Senator Cotton. Do you prefer to see America remain the
world's dominant global superpower?
Mr. Dorsey. I prefer that we continue to help everywhere we
serve and we are pushing towards that, but we need to be
consistent about our terms of service and the reason why. And
the reason why is we also have a right and a responsibility to
protect the privacy of the people on Twitter from constant 24/7
surveillance. And we have other methods to enable any issues
that an intelligence community might see, to subpoena and to
give us a proper legal order, and we will work with them.
Senator Cotton. I have to say I disagree with any
imperative to be consistent between the governments of China
and Russia on the one hand and the government of the United
States on the other hand. Or would you be consistent or even
handed between the government of China and the government of
Taiwan?
Mr. Dorsey. What I meant was a consistency of our terms of
service. And of course there will always be exceptions, but we
want to have those go through due legal process.
Senator Cotton. Let me turn to the actions you've taken
about the 2016 election--both of your platforms--and
specifically one action you haven't taken. You have removed
several accounts as a result of your own investigations and I
think some of this committee's work--and I commend your
companies for that.
One set of accounts that remain on your platforms are
WikiLeaks and Julian Assange. Secretary of State Mike Pompeo,
when he was the director of the CIA, characterized WikiLeaks as
a non-state hostile intelligence service. This committee has
agreed with that assessment now for a couple years in a row,
yet, both WikiLeaks, which propagated some of the leaked emails
in the 2016 election from the Democrats, remain active on both
Facebook and Twitter as does Julian Assange.
Ms. Sandberg, could you explain why Facebook continues to
allow their accounts to be active?
Ms. Sandberg. I'm not going to defend WikiLeaks and I'm not
going to defend the actions of any page or actor on our
platform. WikiLeaks has been public information. It's available
broadly on other media and as such it doesn't violate our terms
of service and it remains up on our site.
Senator Cotton. And Mr. Dorsey.
Mr. Dorsey. We also have not found any violation of our
terms of service, but you know we are open as always to any law
enforcement insight that would indicate a violation of our
terms.
Senator Cotton. Thank you. My time has nearly expired.
Again, I want to commend your companies for making you
available and both of you for appearing. I would urge both of
your companies, or any company like yours, to consider whether
or not they want to be partners in the fight against our
adversaries in places like Beijing and Moscow and Pyongyang and
Tehran, as opposed to evenhanded or neutral arbiters. Thank
you.
Chairman Burr. Senator Reed.
Senator Reed. Well thank you, Mr. Chairman. Let me begin by
thanking you and the Vice Chairman for recognizing my ex
officio colleague Senator John McCain. We are both service
academy graduates, so we don't know any Latin so we had various
translations of ex officio. The one we liked best was real
cool. So you were real cool, Mr. Chairman. Thank you.
Thank you both for being here. You have been organizing,
based on your comments today, very diligently for the 2018
elections and trying to anticipate malign activities that we
saw in 2016.
Have you seen the same type of coherence starting with Ms.
Sandberg, from the Federal Government in terms of your ability
to contact them to work with them?
Ms. Sandberg. We've long had very good relationships with
law enforcement. We've worked closely with DHS and FBI for a
long time. And the FBI's new task force on this has been
particularly helpful.
Senator Reed. Mr. Dorsey, your comment?
Mr. Dorsey. We've also had really strong relationships with
the government. We're always looking for opportunities to
improve our partnership and I think if I were to list them out
it would be a more regular cadence of meetings. It would be
more proactive information about secular trends that they're
seeing, not just on our platform, but other platforms and also
in other channels and communication methods. And, finally, a
consolidation of points to contact--more of a single point of
contact. And we do have that consolidation for the 2018
elections, which we're really happy with.
Senator Reed. Very good. One of the rules is to follow the
money. And you've talked about how you, in terms of political
advertising, have identified the citizenship of their
advertisers but are you able to trace the monies? It's fairly
easy to set up a corporation in the United States, and the
money could all be coming from overseas even from some
pernicious sources. Do you go that far Ms. Sandberg? And then
Mr. Dorsey.
Ms. Sandberg. Sir, you're right that there a lot of ways to
try to game the system and so we are going to keep investing
and trying to get ahead of any tactics our opponents would use,
including that one.
Senator Reed. Mr. Dorsey.
Mr. Dorsey. Sir, we do our best to understand the intent
and where people are located and what's behind them, but this
is where a strong partnership with government comes in. Because
we will not always be able to infer agendas or intent or even
location in some cases.
Senator Reed. In the dialog that you've talked about with
Lauren Forsman, is this one of those topics where you're asking
them for information, or they're asking you and you're trying
to follow the money, or have you seen any of that, or has it
been sort of one of those issues that's just too hard to think
about?
Mr. Dorsey. It's both. We have seen proactive outreach from
the other side.
Senator Reed. But that would be, I think, a critical issue
in terms of governing the behavior campaigns, and I would hope
that you would continue to work, and we would urge our
colleagues in government to work with you, in that regard.
One of the issues, and I think Senator Warner and several
others have brought it up, is the prevalence of bots. I'm not a
technologist, but it seems to me that you could identify a
bot's presence, that you could notify your consumers that 35
percent or 80 percent of these messages have been generated
electronically. Is that feasible? And is that something you're
doing?
Mr. Dorsey. It's a mixed answer right now. We are able to
identify automations and activity coming through our API, and
to Senator Warner's comments, we would be able to label that
with context. But we are not necessarily as easily able to
identify people who might be scripting our website, so making
it look like it's an actual human or even the app--make it look
like an actual human performing these actions. That becomes
much more challenging and unclear.
So in consideration of labeling and context, we need to
make sure that when people see that bot label, that they're
assuming that everything it's not on is human. We need to make
sure that there's a precision and accuracy as we label those
things.
Senator Reed. Wouldn't there be a value in beginning the
labeling process, even with the heavy disclaimer that this
identifies only a fraction of potential fictitious actors?
Mr. Dorsey. Yes, it's definitely an idea that we've been
considering, especially this past year. It's really up to the
implementation at this point.
Senator Reed. Ms. Sandberg, your comments?
Ms. Sandberg. This is one of the ideas I had an opportunity
to discuss with Vice Chairman Warner yesterday in his office
and is in his white paper, and we're committed to working with
you on it.
Senator Reed. Thank you. Let me just ask you a question.
Going forward, I think we're going to come to a major debate
within this country or in the whole world of who owns my data,
which rapidly is becoming me. Is it a company like Facebook? Is
it a company like Twitter? Which raises the question of do you
believe that your users should have the right to control what
you do with their data, either selectively, on an individual
occurrence, or generically, or even simply purge it at some
point? Do you believe that should be----
Ms. Sandberg. Yes, very strongly. It's your information.
You share it with us. If you want to delete it, we delete it.
And if you want to take it with you, we enable you to download
it and take it with you.
Senator Reed. What about for those people who--I think many
people--who in the hustle and bustle of everyday, that's a very
cumbersome process? Shouldn't they be allowed to sort of have a
check that says every two months delete it? Or delete it as
soon as I put it in?
Ms. Sandberg. Yes, and we're working on some of those
tools, and we've improved. We've made it easier to understand
what information we have, how we're getting it, and how we use
it. And we're going to continue to iterate here.
Senator Reed. Mr. Dorsey, the same question.
Mr. Dorsey. We do believe people should have complete
control over their--of their data. Again, Senator Warner
brought up an interesting point earlier, which is--I don't
believe that there's a real understanding of the exchange being
made in terms of people performing activities on these services
and services like Twitter, and how they can actually see that
as an exchange--an exchange of value. And those are things I
would love to think a lot more about, how do we make that more
clear? And I think that goes back to the incentives
conversation.
Senator Reed. Thank you. Thank you, Mr. Chairman.
Chairman Burr. Thank you, Senator Reed, and I thank all the
members for their questions and our panelists for their
answers. I'm going to turn to the Vice Chairman for any last
comments he might have.
Vice Chairman Warner. One, I want to thank you both. I want
to thank you for the spirit you brought to this, some of the
suggestions--your responses to some of the suggestions. I wish
our members were still here, because I think they all performed
extraordinarily well.
I take away from this three or four quick points. One, very
much appreciate, Mr. Dorsey, your acknowledgement that we ought
to move towards--and I guess Ms. Sandberg echoed this as well--
some ability to indicate to users whether they're being
contacted by a machine or a human being, recognizing there's
technical difficulties, and also acknowledging that just
because it's a bot that does not inherently mean it's good or
bad. It just must be a data point that an individual ought to
have as they make determinations going forward.
I also really appreciated, Ms. Sandberg, your notion that
not only should users have access to all of the information
that you or others are collecting, but as we work through to
this--how you monetize that and let users know the value of
their data, I think that increased price transparency--and I
was very grateful at your willingness to at least consider
that, because I think that would go a long way towards making
this exchange better understood by individuals.
Also, and I didn't get a chance to really get into this at
length, but you and I have had this conversation in the past
around data portability. I don't want to make the complete
analogies--an old telecom guy--but when number portability came
around, we got a lot more competition in the wireless industry
and elsewhere. Data portability--I know you make it available
right now--but in an easy, user-friendly format that can move
from platform to platform, I think would be extraordinarily
important in terms of making sure that we continue to have
competition in this space.
And then finally, I also appreciated your comment--I think
we're going to have more and more of these areas where
manipulation may take place that actually incents violence. We
both cited the horrible example of what's happened with the
Rohingya in Myanmar, but I appreciate your comment that you've
said that Facebook ought to have both a moral and legal
obligation if there are sites that are incenting violence and
take those down. Getting from that idea into how we spell that
all out will be a challenge, but I appreciate your willingness
to work with me on it.
So Mr. Chairman, thank you for the fourth hearing on this.
I think it was very, very important, and I hope our committee
will continue to take the lead on these subjects.
Chairman Burr. I thank the Vice Chairman. I would ask both
of you if there are any rules, such as antitrust, FTC
regulations or guidelines, that are obstacles to collaboration
between companies, I hope you'll submit for the record where
those obstacles are so that we can look at the appropriate
steps that we could take as a committee to open those avenues
up.
I want to thank both of you for appearing today and for
your continued efforts to help find a solution to the
challenging problem. This hearing represents the capstone of
the fourth piece of the committee's investigation into Russian
interference in the 2016 elections. So far we've completed our
inquiry into the attempted hack of State elections
infrastructure, the intelligence community assessment on
Russian activities in recent U.S. elections, the Obama
Administration's policy response to those operations.
With your testimony today at this, the fourth hearing we've
held on social media, we heard the top-level perspective on how
to address foreign influence operations on your platforms. When
this committee began its investigation into Russian
interference in the 2016 elections, neither Mark nor I fully
appreciated how easily foreign actors could use social media to
manipulate how Americans form their views.
Like most technology, social media has the capacity to be
used for good as intended, but also to advance agendas of those
bent on manipulation and destruction. Given the amount of
information companies like Google collect on each and every
American, it is also too easy for bad actors to craft a message
that appears tailored just for you.
The Russians undertook a structured influence campaign not
against the American government but against the American
people. Moscow saw the issues that talking heads yell about on
cable news--race, religion, immigration, and sexual
orientation--and they used those to sow discord and to foment
chaos. They leveraged our social media to undermine our
political system as well, but make no mistake, Russia neither
leans left nor right. It simply seeks turmoil. A weak America
is good for Russia.
I think it is also important to highlight that there is a
very human component to all of this. No single algorithm can
fix the problem. Social media is part of our daily lives. It
serves as the family newsletter, a place to share life's
personal joys and sorrows, a way to communicate one's status
during a crisis, and everything in between.
Unfortunately, other states are now using the Russian
playbook, as evidenced by the recently uncovered Iranian
influence operations. We're at a critical inflection point.
Will using social media to sow discord become an acceptable
tool of statecraft? How many copycats will we see before we
take this seriously and find solutions? Your companies must be
at the forefront in combating those issues. You know your
algorithms, your customers, and your data collection
capabilities better than any government entity does--or should.
Still, the burden is not entirely on your shoulders.
Government, civil society, and the public will partner with
you.
I'd like to take just a moment to thank our staff. They
have worked diligently to uncover the scope of the problem.
Their research has been thorough. Their efforts are seamlessly
bipartisan and their drive to defend the public against foreign
influence should make Americans watching today proud.
There is no clear and easy path forward. We understand the
problem and it is a First Amendment issue. We cannot regulate
around the First Amendment, but we also cannot ignore the
challenge. I am confident that working together we can find a
solution and a path forward that will only make us stronger,
more connected, more prepared to face down those who seek to
weaken our democracy.
For your participation in being part of the solution, we
thank you immensely today.
This hearing is now adjourned.
[Whereupon, at 12:11 p.m., the hearing was adjourned.]
Supplemental Material
[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]
[all]