Mark R. Warner is a Democratic U.S. senator representing Virginia. He served as the Governor of Virginia from 2002 to 2006. This is the transcript of an interview with Frontline’s James Jacoby conducted on May 24, 2018. It has been edited in parts for clarity and length.
How significant was the role of Facebook in the Russian influence campaign during the 2016 election?
We’re in the middle of 2018, and we still don’t have the full answer of how extensive the Russians used Facebook. We do know that Facebook and Facebook-related posts touched about 150 million Americans that were posts that originated either through Russian fake accounts or through paid advertising from the Russians.
We originally thought that the paid advertising was the main focus and a little disappointed at first. Facebook was a little slow on the draw and originally only identified those ads that were paid for with rubles. They didn’t even look at those same accounts to see if they might have bought ads with dollars or euros or some other denomination. But the paid advertising was really a relatively small piece of the overall problem. A much bigger problem was the ability for someone to say they were James in Washington, D.C., but it was actually Boris in St. Petersburg creating a fake persona that would generate followers. Then once those followers are generated, often on a topic that had no relationship to anything like politics – it might have been gardening; it might have been football; it might have been some other sports agenda – then they would seed it with the fake information and the false news and the political content.
Those fake accounts – and they numbered around 470 that were identified out of a single enterprise in St. Petersburg called the Internet Research Agency, an entity owned by one of the Russian oligarchs – was the extent of the outreach that we’ve seen, although increasingly the intelligence community believes there were other, more directly state-sponsored activities by some of the Russian spy agencies that we still now, almost 20 months later, still haven’t fully identified.
When did the intelligence community become aware of the fact that there was a Russian influence campaign being waged on social media sites like Facebook?
I think we had seen Russian disinformation/misinformation campaigns in Ukraine and the Baltic nations. But in many ways, the American intelligence community was caught pretty flat-footed. I don’t think we fully appreciated what they were doing until after the election, and I will give Facebook – perhaps they were not even fully aware as well. What I recall is as early as November of ’16 and early December of ’16, I raised this issue that it appeared that there had been a lot of Russian activity, not just on Facebook, but Facebook, Twitter, YouTube, other platforms, and the initial reaction from the Facebook leadership was: “Oh, that’s crazy. People must not understand our platform if they thought Russians could misabuse the Facebook platform.” It was very shortly after that that Mr. Zuckerberg and the Facebook leadership had to reverse its course and acknowledge that this was a problem, and by the time of the French presidential elections, which were the spring of 2017, Facebook had actively engaged with the French officials and took down 30,000 accounts prior to the French elections that were connected with Russian activity. Again, one of the reasons why I feel like we may not still know the full extent of the interference in the American election [is] because just on types of scale – 30,000 accounts taken down in France, 470 in America – in my mind, I think the Russians were a little more active than what we discovered so far.
U.S. Intelligence And Disinformation
You say “flat-footed”; you say that the intelligence community was caught flat-footed. It was known for years that the Russians had launched disinformation campaigns. How was this not an intelligence failure?
Well, part of this is that this strangely falls between the scenes. If somebody originates a fake post in St. Petersburg, Russia, and it appears on your phone in Washington, D.C., or Dallas, Texas, on one hand, the NSA or the CIA might be looking into the activity that takes place in Russia, but when it appears on an American’s individual device, that falls on the side of the FBI or the Department of Homeland Security. So this divide between international activities and domestic activities and the real desire – and I think sometimes the American public believes that our intelligence agencies have a lot more exposure on Americans’ personal information than they do. It falls onto the American individual’s device or touches an American person. It really is thrown over to the FBI, so that falling between the cracks a little bit, originating in one country, appearing in another’s, was one of those scenes. While we’ve seen misinformation/disinformation campaigns waged in other nations, we’ve not seen it to the extent of this many false personas that had been created and literally had created, in cases thousands and in aggregate millions of followers with these fake identities.
Were the NSA [National Security Agency] and the CIA communicating with the FBI and Department of Homeland Security about the potential threat here?
I think information was thrown over the transom from those spy agencies that focus on international activities, threw it over to the domestic agencies. But we’ve still got room for improvement. I mean, we created after 9/11 something called the Director of National Intelligence that’s supposed to increase this communication between our literally 17 different intelligence agencies. We’re better than we were, but I think 2016 showed particularly in this realm of social media that we were behind the curve.
And in terms of coordination and threat assessment, was anyone from the intelligence community or the FBI or DHS, Department of Homeland Security, communicating the possible threat to Facebook and the other companies in the lead-up to the election?
I think there was communication on an ongoing basis, but the willingness of Facebook and for that matter most of the social media platform companies and for that matter much of the valley to interact with our intelligence community, those connections have been frayed, partially frayed because of the revelations that came out of the Snowden activities, partially frayed because of some of the debates about encryption and back doors in devices. Now, I think that that kind of communication needs to improve, because if individuals in increasing numbers don’t believe they can rely upon the identity of individuals that might pop up on their Facebook feed or an account that would pop up or for that matter when Americans get over half of their news now from Facebook newsfeeds, if that information is suspect or fake news or being originated from a foreign source but trying to act like it’s an American-sourced piece of information, people start to lose credibility with their platform, so it’s actually in Facebook’s own long-term economic interest to work with us.
The one thing I think I’ve tried to appeal to Mr. Zuckerberg and others is while I had 20 years’ experience in the technology business before I came to Congress, many of my colleagues, as was exhibited by the recent testimony, really don’t get this stuff very well. If they leave it just to the politicians to try to come up with a solution set, chances are we’ll mess it up, so this will require much greater collaboration than we’ve seen to date.
But let me ask you specifically: Are you aware of any conversations or any communication between either the intelligence community or Department of Homeland Security, the FBI and Facebook or the other platforms before the election about the potential threat from foreign adversaries gaming their platforms?
There were communications. Whether there was specific threat analysis in terms of the potential for Russia to abuse these social media platforms, I’m not aware.
Congress And Facebook
So what was it? What was it that launched your investigation initially?
Well, what launched my interest was there were a number of rumors. I had some great staff who were starting to educate me about how the social media platforms could be abused, and there seemed to be strange – not only strange voting patterns, but also a great deal of political activity that seemed to take place in places where it appeared somebody knew more than frankly the Democratic candidate actually knew about where the election was headed.
When I raised this issue, I was pretty disappointed with the initial pushback from Facebook, saying, in effect, “This is crazy; Warner didn’t understand what he’s talking about,” when in actuality I believe by after the election there was enough buzz that I would have hoped that Facebook would have started their own investigation of that point to realize that there was something a little fishy both in terms of paid political ads originating out of Russia, paid for in rubles, and then this volume of traffic that was masquerading in identities and obviously directly intervening.
Then the whole intelligence community finally connected all the dots with an intelligence community assessment, and it came out in January of 2017 that [there] was – recently, my committee reaffirmed its findings, and those findings were that Russia had massively intervened; that they had hacked into both political parties but leaked information only to help Mr. Trump and hurt Hillary Clinton; that they had scanned or touched 21 of our state’s voting systems; and that they had massively used social media in ways that frankly, at least in America and I would say in the world up to 2016, was unprecedented.
Facebook’s Response To The 2016 Election
The intelligence community is saying that in January 2017, right, just a few months after the election, yet Facebook’s response to you is that this is crazy?
Well, Facebook’s response to me was prior to that intelligence community assessment, and I first raised these concerns in November of ’16. So literally two months later, the intelligence community assessment came out. Facebook was starting to change its position. I was out at Facebook in the spring, and they were acknowledging the problem; they were talking proactively about how active they’d been involved in the French elections and working with the French authorities. But it was literally months later, by the summer of 2017, before we started to get actual numbers of accounts, the number of examples of those accounts, the literally thousands of individuals that followed these accounts. And we only received the actual documentation data and copies of the ads literally late, late summer of ’17.
I want to come to that, but I’m just curious about the initial response, because here you are, a U.S. senator on the Intelligence Committee who is going to Facebook, also a former tech person, and saying to them, “Look, you guys need to start looking into what role there was that you played during this election.” Were you surprised by the recalcitrance?
I was disappointed. And I had raised these issues. I had meetings in early spring. We then had the French elections. I went out to Facebook in late spring. There is, and not just with Facebook, but there is this sense from some of my friends in the valley: They are smarter than everyone else and that somehow Washington policymakers don’t understand anything. I think for a while, remember, these social media companies have created a wonderful whole new means of communication, a whole new ability for folks particularly – you see some of the positive aspects from beginnings in Arab Spring – a whole new ecosystem. And I think there was this sense that they can only do good, when in effect, there’s a lot of good that’s come out of the social media enterprises, but with every good there is a dark underbelly, and that dark underbelly, I think they were reluctant to have exposed, because at the end of the day, when people start losing faith in how the information is relayed and how their personal information might be being abused and misused, that hurts their business model.
Who was it specifically that told you you were crazy for looking into this?
Well, the initial comments actually came from a tweet from Mark Zuckerberg, and I’ve reminded him of that a few times since that time.
And has he apologized?
I think he’s acknowledged that they made mistakes; they were unaware of how serious this problem is.
You really believe that they were unaware of how serious this problem is given ample evidence that in the past, their platform had been gamed elsewhere?
Listen, I’m going to give them the benefit of the doubt, but for a company that makes its revenue based upon knowing as much as possible about their user community and then selling that ability to advertisers to target that user community, it did stretch my imagination a little bit that they weren’t further aware.
In April, Alex Stamos, the head of security at Facebook, came out with a post, and they announced basically that there was – they didn’t mention Russia. They said a few “bad actors” had tried to do some things on the platform. Do you remember what your response was to that when it came out?
My reaction was they still don’t get it. They still are not willing to come clean. They’re not willing to acknowledge that a foreign nation intervened in our democratic process, used our open society and our open rules to try to basically mess with the results of an election. I was very disappointed by the initial post.
What kind of good corporate citizen is this, right? I mean, these are very serious allegations about interference with our elections using their platform.
Well, I would also say that there was similar allegations made against Twitter, against components of Google. If this was a mindset that spread across all the social media platforms, I don’t think we can single out Facebook alone. There was, in a sense, collective amnesia about perhaps what happened, and I think an unwillingness to deal with the future potential misuse and a sense, I think, sometimes from some of the folks in the valley, that if they ignore Washington, it will just go away. This was just the beginnings of what now has blown into I think a full-fledged crisis, not only in this country, but we’ve now seen, we had recent indication that Russian social media activity has taken place in all 29 NATO nations. Now this is a worldwide phenomenon, and the cost-benefit relationship for the Russians – I like to say if you look at all of the cost of the Russian intervention in American elections, in the French election, now there’s evidence again in the Brexit vote in the U.K., you add that all altogether, that’s less than the cost of one new F-35 airplane.
This is a pretty good return on the dollar, return on the ruble for the Russians, to use these tools, and that playbook now was out there for China, for other nations, frankly for cybercriminals, and we’re still trying to get our arms around the idea of fake identity. We’re in a sense chasing 2016 technology. The next wave of this manipulation will come in what’s called deepfake technology, where a viewer would see on their handheld device an actual live video stream with what appears to be my face and voice, or for that matter a business leader’s face and voice, or for that matter Mark Zuckerberg’s face and voice, and that face and voice has no relationship to that actual human being.
It’s a total fake. And when that starts to roll out, the effect it will have not only on elections but the effect it could have if they – if the Fed chair’s image was presented, how it could roil markets. The ability to compete between Coke versus Pepsi in terms of normal corporate war and trade war has huge ramifications, and that’s why these companies need to realize there’s got to be some guardrails.
You go out in May to visit with Facebook. Who did you visit with Facebook in May?
I saw senior leadership. I didn’t see Mr. Zuckerberg, but I saw all the other senior leadership.
Did you see Sheryl Sandberg?
I saw the senior leadership.
Did they give you sufficient time?
We had a good session, but I still felt that [despite] the amount of resources and focus on this problem that it wasn’t getting appropriate attention. I think I and a few others started to ratchet up, and I think increasingly more of them, the media, started to focus attention. I think the amount of resources that Facebook put behind trying to investigate what happened, looking backward, dramatically increased. We also put forward what we thought was the lowest hanging fruit in terms of a legislative solution, at least the requirement for foreign-based political ads to have the same disclosure requirements, and for all political ads to have the same disclosure requirements on social media that they have on television or radio or newsprint. Facebook and Twitter, for example, have embraced most of those principles, but, as I said, the paid political advertising piece is really only the tip of the iceberg.
In September, Facebook releases some numbers about their ads. What was your response, and what was your reaction at the time to that release?
Well, by the time [of] the release on the ads, I realized that the ads [were] in a sense just the tip of the iceberg; that the ads were relevant but was not the scale. They identified 150,000 roughly paid advertising in social media that still can touch a lot of individuals and make a lot of impressions when really that was not where the action was. Where the action was was the creation of fake accounts, many of them geared politically, but many of them geared simply to cause dissension among Americans.
Basically this is almost a year after the election, a year after the allegations surfaced of Russian activity on Facebook. Did you think that that was the best that they could do with a year of investigating this issue, was 100-some-odd-thousand dollars in an ad buy and a few thousand ads?
I did not think their response was adequate, and I made that known on a repeated basis.
Do you think that they were not disclosing the full truth at that point?
I, again, hope that they didn’t have the full truth at that point, that they were still going back and doing the screening of where these accounts originated. Attribution in the cyberworld is always a challenge, particularly as accounts go from one hop to another hop, particularly with different routers. As a former telecom guy, I can get pretty deep in the weeds here, but I didn’t think they were throwing the resources at trying to ferret out the number of accounts and the amount of abuse and the amount of misinformation that was out there.
Despite the seriousness of the allegations.
Despite the seriousness of the allegations. Now again, they continue to get better. The circumstance, though, was there was no social media actor that really stepped up right away. We had the same lack of response from Twitter. Matter of fact, originally all of Twitter’s work in this field was actually derivative of the Facebook work. There was not that much focus now from Google in terms of their YouTube portion of their platform. There’s still been less than full response from Reddit. So a number of these platforms, I don’t think took it seriously enough, and increasingly we found that the bad guys have found more and more sophisticated ways to mask identities, to create new tools. This is going to be a running battle on an ongoing basis, and if the social media companies don’t step up, you’ve already seen the European Union now take action with a new set of privacy rules and data portability rules, the ability to kind of take yourself out of the game. There are a set of policy ideas that I have. I don’t want to kneecap great American innovative companies, but I also want to make sure that I think individuals have a right to know, for example, if they’re being contacted or manipulated by an actual human being or by an automated account or a bot. I think we need to do – there needs to be some at least minimum due diligence around identity or at least origination, geographic origination of a post. That gets more and more complicated as we move into more information flowing through the cloud.
But these are areas where we need to lean in, or you’ll see policymakers start to say that perhaps these social media companies ought to be viewed as media companies, and they have the same kind of content curation responsibility that your network has. We’ve already seen small movements, legal movements in that area around prohibitions on child pornography, more recently legislation that’s been signed that prohibits sex trafficking, so this whole area around content which is called Section 230 exemption. That’s another. I’m not saying I’m there at that, that that’s the right solution set, but there are a whole variety of models that we could get to from a policy prescription including some that would end up disaggregating between the user and the social media company, where there might be some trusted intermediary that would help provide some of that screening or help guard your data if you as the individual own that data, and in effect that the social media companies could never take possession.
Time is limited, and I just want to make it through our timeline, because I think it’s really important for a public accounting of the timeline. In September, they tell us that it was an ad buy, $100,000 in ads, 2,000 ads. Independent researchers find very quickly that there was much more of reach than that ad buy. You know tech. They know their business model; they know it’s not just advertising; they know that organic content reaches far more people. Do you think that they were being deceptive by just releasing the ads?
I was disappointed that it took them a couple of months more before we started getting where the real action was, the accounts and the type of accounts and some of the actions. There was one example we cited in one of the public hearings that we had with the lawyers from the three social media companies back in November of ’17, where one account was set up to try to rally the Muslim community in Texas. Another was an attempt to rally the right wing in Texas. They created an event, a protest, with both sides protesting against each other at a mosque in Houston in October of 2016. But for the good work of the Houston police you could have had the kind of horrible activity take place then and there that I saw unfortunately take place in Charlottesville in my state last year, so the real human consequences of some of this abuse. We’ve been very lucky that it hasn’t actually cost people’s lives.
… Your November hearing, the companies send their general counsels, their lawyers, not their chief executives. What did you make of that?
I thought their sending of their lawyers was a pretty lame action, and it didn’t meet our full requirements. I believe we still need back the CEO of Facebook, Twitter, Google, and potentially others to come testify, now.
How were their answers about that? How were the lawyers’ answers?
The lawyer’s answers were great lawyer answers, which basically didn’t fully address all the timeline, did indicate that they were making changes in policy, did indicate that they’d, for example, come to support the tenets of the Honest Ad[s] Act, our bipartisan legislation about at least disclosing the source of political advertising. It was again, incrementalism, but as we see this problem widen on a worldwide scope, and when we see them coming on top of the potential misuse by Cambridge Analytica of scraped Facebook personal data, that many, many people didn’t know that their information was being used by this entity that was employed by the Trump campaign to target voters, the problems continue to mushroom. So I believe they’ll be back.
I want them back, and I want to do it in the context of where we can lay out a series of policy reforms and let them respond to those reforms.
… The Cambridge Analytica revelations, how did that strike you at the time? I mean, here’s another thing on top of what you already –
Cambridge Analytica had a bit of a sketchy reputation in terms of working in a variety of countries in a pretty manipulative fashion, and it seemed they had a lot more results, and it was fascinating to me that they claimed that they were the brains behind the Trump victory in terms of their targeting. After the election, the Trump campaign started to distance themselves from the claims of Cambridge Analytica, but what had happened was Cambridge Analytica had actually solicited individuals to in effect do a survey, receive a small stipend for doing that survey. But what was happening is they would do that survey, [and] Cambridge Analytica would come in and scrape all of the individuals and in fact friends that in no way had given up that personal information. I think Facebook should have been able to ascertain that at a much earlier basis.
Now they’ve changed in terms of the ability for third-party entities to get at that kind of access. But let’s face it: At the end of the day, the Facebook model is based on the notion of collecting as much information as possible about their users and then using that information to lure advertisers to buy advertising on the Facebook platform. There are a lot of the algorithms that Facebook I believe used, particularly when we’re starting now to move into the news area, where we’ve seen, and there have been folks who used to work at Facebook and at Google and others who, independent researchers make these claims that regardless of which way you lean politically, you read one story, and you in effect get a little dopamine hit, and the next story you read is more outrageous, simply reinforcing your view. I believe that those techniques are still very troubling in their basic business model.
This is a news source for nearly half of the voting public, and –
Actually over half of the voting public at this point.
Regulating Big Data
Should it be regulated as a utility?
Well, I’m going to lay out in the coming weeks a series of policy options to try to start this debate, and it will range from questions around content to questions about the ability to know whether you were being contacted by an individual or a bot, questions around identity – although, again, in many nations, there are legitimate reasons why someone may want to hide from a government in terms of true identity. That’s problematic. As a former guy that did a lot of work in telecom, it used to be really hard to move from one telecom provider to another. There’s something called number portability. You could have the same kind of notion around data portability. Right now Facebook says you can leave your platform, but it’s really hard to migrate all your past information and videos and friends from one platform to another. There’s this area as well of making sure that you potentially own your own data, and you might have in effect an intermediary be that guardian of your data and only let it out to certain platforms for certain uses. So there’s a wide range of policy options, and they don’t fall kind of liberal/conservative; they really fall future/past. What my appeal to all of the social media companies are is that they are relying in many ways on their users’ trust, and that trust over the last year with the fake news, with the abuse by the Russians, with the abuse by Cambridge Analytica that trust is eroding for a lot of Facebook users.
Facebook has a pretty big lobbying operation as do all of the Internet companies that give heavily to both parties. There’s a revolving door of people in government going to those companies including former chiefs, chief of staff of yours, who is now at Facebook. Are you going to get any political traction to be able to do something about this given the amount of power that those companies have right now?
I put the first legislation forward called the Honest Ads Act, and we’ve gotten support from social media companies. but that’s just the beginning.
Was that support all along?
No, that’s support again. I’ve never known any enterprise, not just social media companies, to greet any kind of regulatory approach with open arms on the first day. You’re right: There are plenty of lobbyists, but at the end of the day, at least this senator is going to continue to make the case. I want American innovation to continue. But there has to be some guardrails, and simply disclosure alone around political advertising will not be enough. But one of the challenges is to candidly educate my colleagues. If there was one thing that came clear, and I say this respectfully about the Zuckerberg hearing, was that an awful lot of senators don’t understand the basic tenets of how social media works, what the business model is, what the technology really involves. And my hope is that I can lay out a series of policy options and build a bipartisan coalition that will say, yes, we want innovation; yes, we want the benefits of having these kinds of connections. But there’s got to be a greater ability for users to have trust in their content.
Can we just ask you off the record, do you feel like they have turned everything over to your committee that they – did they give you everything you need?
They say they have, but I still would say this, on the record, that –
You would say this on the record?
Yeah. I still have questions. In 2018, we’ve identified the Russian activity that originated out of the Internet Research Agency in St. Petersburg. I still have an open question whether the Russian spy services had other activity that was used to interfere in our elections, and I don’t know the answer there. Unfortunately I don’t have a definitive answer from our own intelligence community still whether there was that additional posting, additional fake information, and I don’t have an answer from Facebook. That still disturbs me on a going-forward basis if we’re going to make sure that in 2018 and 2020 this doesn’t happen again at a scale and scope exponentially larger than what happened in 2016.
originally posted on pbs.org
