One of the Largest Disinformation Campaigns Ever Conducted

CHRISTIANE AMANPOUR: Now, the money spent in U.S. presidential elections, as we have been saying, is mind-boggling to so many people. And President Trump himself says he is ready to spend a billion dollars if it helps him stay in the White House. “Atlantic” journalist McKay Coppins created a new Facebook account so that he could follow MAGA-related pages, join groups and receive messages from Trump supporters. And he told hurry our Hari Sreenivasan what he learned about the campaign’s aim to spread disinformation, discredit journalists, and even dismantle mainstream media.

HARI SREENIVASAN: McKay, thanks for joining us. So, I want to start with a couple of paragraphs in your piece that are the worst-case scenario after all of your research and diving into this story. “On Election Day, anonymous text messages direct voters to the wrong locations or maybe even circulate rumors of security threats. Deepfakes of the Democratic nominee using racial slurs crop up faster than social media platforms can remove them. As news outlets scramble to correct the inaccuracies, hordes of Twitter bots respond by smearing and threatening reporters. “Meanwhile, the Trump campaign has spent the final days of the race pumping out Facebook ads at such a high rate that no one can keep track of what they’re injecting into the bloodstream. “After the first round of exit polls is released, a mysteriously sourced video surfaces purporting to show undocumented immigrants at the ballot box. Trump begins retreating rumors of voter fraud and suggests that Immigration and Customs Enforcement officers should be dispatched to polling stations. “‘Are illegals stealing the election?’ reads the FOX News Chyron. ‘Are Russians behind the false videos?’ demands MSNBC.” If it was 10 years ago, I’d say that sounds like science fiction. Why is this so plausible to you after what you have been looking into?

MCKAY COPPINS, “THE ATLANTIC”: Well, what I found over the course of the last several months as I was researching and reporting this, this article, is that there is a new and sophisticated suite of tools that are being deployed right now in the United States by political operatives and various political coalition’s to actively advance disinformation that helps their candidate. In the case of this story and what I was focusing on, the coalition to reelect the president — that includes the campaign, partisan media, pro- Trump political operatives — all of them are working in concert to advance false narratives, fan conspiracy theories, and basically to confuse and disorient the electorate enough that he is able to win reelection. And I put that worst-case scenario in the piece in part to kind of illustrate what could happen on Election Day. Some of the experts I talked to said that doesn’t go far enough, that, given the scale of the tools that are available to political campaigns and the brazenness with which people on both sides, but especially those in the president’s orbit, have shown in kind of advancing misinformation, it could get even worse than that.

SREENIVASAN: I can see politicians on the other side of the aisle looking at your piece and saying, hey, how is this different than what every president tries to do? They have the apparatus of the Democratic Party at their disposal. They have the power of the bully pulpit there. They have the ability to do these things. So what’s so different about this? And why are you calling what the Republicans are doing disinformation vs. what has always happened in politics, is that people try to protect the power that they have by any means necessary?

COPPINS: It’s a fair point. And I should say that every presidential campaign I have covered and certainly a lot more campaigns before that have featured a fair amount of partisan spin and half-truths and outright lies, right? We know that political candidates say things that aren’t true on the campaign trail all the time. What’s different is both the type and the scale, I think. First of all, the tools that are available now make it a lot easier for those lies and spin and half-truths to travel around the Internet. And a lot of voters on both sides of the aisle are so firmly ensconced in their own information bubbles, that they’re more susceptible than ever to this kind of disinformation, without ever being exposed to fact-checks or information that might challenge the authority of what they already believe. The other thing is that we have not had a president, at least in a couple generations, this willing to say things that are flatly untrue. I should say, I’m not talking about matters of ideological opinion or viewpoint, right? And there are all kinds of conservative arguments and liberal arguments that I think are completely reasonable. And I — in my perfect world, 2020 would be a year where we have a big ideological debate in this country. But that debate is very difficult to have when there isn’t a common set of facts that most Americans are working from. And what the Trump campaign is actively trying to do — and I saw this firsthand as I was reporting this story — is to make it so that there is no common set of facts, that people are so confused and kind of disoriented by the onslaught of information and propaganda, that they just throw up their hands and say, there’s — there’s no way to separate out what’s true and what’s not, so I’m not even going to bother.

SREENIVASAN: Break down for us, what is the connection between, say, for example, when the president tweets something, what happens and how does that loop work?

COPPINS: Well, as part of the research for this story, I actually created a separate Facebook account from the one that I typically use and subscribed to President Trump’s page, that of his reelection campaign, and then kind of followed the Facebook algorithm as it suggested various other right-wing accounts for me to follow. And what happened was, I created a Facebook feed for myself that was — that was completely filled with the messaging and content that the president and his allies are pushing out every day. And I have to say, it’s a lot more sophisticated than I think a lot of people realize. Some of it is just sort of parroting what the president says on Twitter or what he will say at a press conference. But a lot of it is actually very slickly edited videos. For example, during the impeachment proceedings, there were days when I would be watching the impeachment hearing live on TV. I would take stock of the witnesses and the evidence that was presented and sort of draw my own conclusions. And then, later that day, I would look at this Facebook feed and see a video that the Trump campaign had created that actually took out-of-context clips from the same hearing that I had seen, and slapped them together in a way that made it seem like the hearing had exonerated the president or that it had been some sort of major triumph for the president. And I have to say, even as a journalist who had gone into this with my eyes pretty wide open, I became disoriented by it. I — there were moments where I would see one of these videos and say, wait a second, is that what happened? Did I misunderstand what I saw in the hearings? And I could see how easily people of good faith who were not trying to just see information that they agreed with, but they could — they could become confused and disoriented by the constant onslaught of information and basically decide that it wasn’t worth trying to sort out fact from fiction and either disengage or just become disillusioned.

SREENIVASAN: How did Brad Parscale, the person who is Trump’s kind of digital chief, how did he innovate around this? What did he do around the 2016 election that was so dominant? And what’s happening now?

COPPINS: So, in 2016, the Trump campaign, after Trump won the nomination, they kind of looked at the landscape, realized that they were outgunned by the Clinton campaign, did not have the same war chest that Clinton had, and didn’t have the ability to advertise on TV the way that the Clinton campaign was. And what happened was Brad Parscale, who at the time was the digital director of the Trump campaign, decided to go heavy on Facebook and Google advertising, basically convince the campaign that we can get more bang for our buck by advertising primarily on these digital platforms. And he made use of a technique called microtargeting that maybe some of your viewers have heard about, but don’t know that much about. Basically, what it is that you take the electorate, if you’re the Trump campaign or any other campaign, and you slice it into very narrow, distinct, specific niches. And then you can tailor ads directly for those niches. And so what the Trump campaign did kind of masterfully was that they acquired a ton of data on voters in the United States, and would break them into very small groups. So, for example, if they wanted to serve an ad about defunding Planned Parenthood, they could serve it directly to 800 pro-life Roman Catholic women in Dubuque, Iowa, and know that they would get a pretty positive result. Now, they didn’t invent microtargeting. It had been used by the Obama campaign in 2012. The Hillary Clinton campaign used it a bit, but the Trump campaign was different, both in how much it used microtargeting and how many ads it created. There was a five-month period in the 2016 election where the Trump campaign placed 5.9 million ads on Facebook, while the Clinton campaign placed just 66,000. So that gives you a sense of the scale. The other thing that they did was, they were willing to use microtargeting in ways that were kind of outside of political norms. So, for example, in the final weeks of the race, they microtargeted ads directly to black voters in Florida that said, Hillary Clinton believes African-Americans are super predators. That was drawing from the controversial quote that Hillary Clinton had given in the ’90s talking about gangs, and kind of stretching the bounds of what she had actually said, but they served it to black voters in Florida. And there was a Trump campaign official quoted at the time who said that it was not really designed to win black voters over to Trump. Instead, it was designed to depress black turnout in Florida, basically to disillusion voters.

SREENIVASAN: OK. So if there are going to be more Facebook ads in 2020, how consequential is the decision by Facebook as a platform not to fact-check anything, to basically not hold the political ads to the same standard that they would hold if you and I were just selling a product?

COPPINS: Yes, and that’s a key point. I’m glad you brought that up, because commercial advertising on Facebook is held to basic standards of truth in advertising. You are not allowed to run an ad that just blatantly lies about your product. Facebook will pull the ad if you do that. That’s not true for politicians. Mark Zuckerberg actually announced last year when he was under a lot of pressure over this question of how misinformation was spreading on his platform, he said, look, we’re going to crack down on some of it, but we believe that we shouldn’t be in charge of arbitrating political speech. So, if a politician decides to run an ad that’s false, we’re going to allow that to happen and leave it up to journalists and watchdog groups and opposition parties to decide whether the ads are true. Now, to a certain extent, I sympathize with that, because I don’t think that Silicon Valley companies should be the final word on what kind of speech is allowed in our political arena. But, at the same time, when you’re looking at the amount of information that is available on Facebook, the relatively scarce resources that journalists like me have, and the relatively short — relatively small amount of time we have to sort through all of the ads that are being posted, you start to see the problem here. So I think it is a consequential policy decision that Facebook made. And I think that it basically incentivizes politicians to stretch the truth and lie, because it’ll be very hard for them to be held accountable. And it’ll be easy for them to take advantage of the political benefits of lying on Facebook, with relatively little downside.

SREENIVASAN: So one of the things you also point out is that, in addition — I mean, if two or three cycles ago, we were talking about robo-calls, and then maybe a couple of cycles ago, we were talking about e-mails, and even if last year we were talking about Facebook ads, that text messaging is going to be an important part of this cycle. How?

COPPINS: Right. So, until pretty recently, for a campaign to include you in a mass text, you had to opt in. You had to give your phone number to that campaign and say, I’m willing to accept texts from you. What’s happened is that, in the last few years, peer-to-peer texting apps have been created, which basically enable campaigns to hire staffers or even give volunteers this job and have them sit down and just literally click send over and over and over again and send hundreds of messages in the space of an hour. And to the FCC, that kind of texting is not considered mass texting. It’s considered one-on-one texting, which basically has meant that political campaigns have been on this mad scramble in the last few years to suck up as many cell phone numbers as possible, so they can then send messages to them. Because of the advent of these peer-to-peer texting apps, and because there’s relatively little federal regulation around them, it’s entirely possible that, later on in the race, as we enter a new stage of this election, you could see campaigns or outside political groups using these peer-to-peer texting apps to spread disinformation, to confuse people about where they’re supposed to vote, or when the election is, to run false flag operations tagging opposing candidates with positions that they don’t actually hold. There are a lot of experts I have talked to who say this can be a big problem. They believe that regulators will eventually catch up to it, but that, at least in this election cycle, you should be a little wary of any kind of weird political text message you get that you didn’t sign up for.

SREENIVASAN: There have been a lot of stories over the last couple of years — actually, not a lot of stories. I should say there have been a few stories about a lot of different Web sites that look like they are coming from a specific community, but they’re actually not. They’re fronts. And it’s from a centralized place. How does that work?

COPPINS: Yes. So, the last decade or so has been brutal for local journalism. Local newspapers have been shutting down across the country. Something that’s emerged to fill that vacuum, but in a very different way, have been these sort of faux local news sites. I call them Potemkin local news sites, where, if you look at them, they have kind of innocuous names, like The Kalamazoo Times or The Arizona Monitor. They sort of look like local news Web sites. You can scroll through them. They have coverage of local schools or whatever. But if you look closer, you will see that they don’t have local addresses. They often don’t have mastheads or bylines. It’s not clear who’s behind them. A lot of them are actually owned and operated by Locality Labs, which is a company owned by a conservative activist. Others are run by lobbying groups or even local Republican parties. And I spoke to one political strategist who told me how they’re often used, which is that a candidate who wants to get a certain story placed about an opponent, and the local journalists won’t actually take it, they can actually pay a third party to have their desired headline placed in one of these Web sites. And the average reader who’s just kind of scrolling through Facebook and comes across this will have no idea that the Web site has a political agenda. They will think it’s just a normal news story. But that’s what actually makes these valuable.

SREENIVASAN: Is there such a movement at this scale on the left? I don’t want to paint a false equivalence, but what has the left been doing?

COPPINS: This is kind of an open debate among Democratic strategists right now. So I would say that it’s much more sophisticated, much more advanced and much more, for lack of a better word, brazen on the right, but there is another information ecosystem on the last where there are Democrats openly kind of discussing whether they need to co-opt the same tactics that the president is in order to beat him in 2020. What I would say is that we won’t have a really clear idea of how far or how — they’re willing to go until we have a Democratic nominee. But my concern, really, in writing this piece wasn’t with the immediate horse race or who wins this election, but what happens to our information ecosystem going forward, because I do think, if we get into a situation where both major political coalitions in this country decide that the only way to win is to actively engage in propaganda and disinformation and to kind of trick voters into voting for them, then we have a real problem with the health of our democracy that will expand — extend well beyond 2020.

SREENIVASAN: McKay Coppins, thanks for joining us.

COPPINS: Thank you.