Facebook: Tactics by those trying to manipulate U.S. elections are changing
Nathaniel Gleicher, Facebook Head of Security Policy, joins Yahoo Finance to talk about what Facebook is doing to inform voters on rules and regulations ahead of the 2020 presidential elections and how the social media giant is identifying and handling misinformation on its platform.
Video Transcript
- The reason I bring up Facebook is they've announced a new program to potentially safeguard the news and information that is shared on the platform. And to help us understand what this is about, we welcome into the stream, Nathaniel Gleicher. He is Facebook's head of security. Thank you for joining us today.
NATHANIEL GLEICHER: Thanks for having me.
- So let me start with this, and I'm going to use an example that reared its head yesterday afternoon. We have the President of the United States and members of the administration refusing to knock down what is clearly a false bit of news that Kamala Harris may not be eligible to be on the ticket. We know she was born in Oakland. And yet birtherism has reared its head. What does Facebook do when people start sharing articles about that which are clearly inaccurate?
NATHANIEL GLEICHER: Whenever there's a piece of content on our platform that a third party fact checker determines is false-- we work with a broad network of independent fact checkers, for just this purpose-- we will label that content very prominently. For example, if it's a video, there's actually a label across the video that you have to click through before you see the video. And then we'll also downrank that content, so far fewer people will see it.
What that means is that no one stumbles across these videos or this content once it's been labeled as false. And if someone is hunting for it, which as we know on the internet they will find that content somewhere if they're looking for it-- when they do find it on our platform, it has a clear label on it indicating that it's false, and a link to a fact checker who's analyzed what's being asserted in that video.
- So who are the broad fact checkers? You hire them, you pick them. How do you screen them and what do you do to stay on top of making sure that they're doing the job you want?
NATHANIEL GLEICHER: So we work with independent fact checking organizations, all of whom are certified through the--
- Who are they?
NATHANIEL GLEICHER: All of whom are certified through the Pointer System. There's a wide range. So Associated Press, Agence France-Presse, a broad range of fact checkers, all of whom are listed publicly. And every single one is certified by this independent organization. So we use that framework to ensure that we're getting neutral and independent-- and independent fact checkers to do this assessment.
- Nathaniel, I think we can all anecdotally say it hasn't worked terribly well. And I won't-- and I won't limit that just to Facebook, by the way. I think in general, we see misinformation spreading far and wide despite what the social media networks have said or are their best efforts. So what can you do differently, I guess, is the question to shut this down and in more decided fashion? Because I think all of us have seen this stuff, even though you're taking these measures that you're talking about.
NATHANIEL GLEICHER: And I think the question of misinformation and public debate today-- it's a very big challenge beyond the social media platforms. I know-- I'm sure you struggle too with how you cover bald-faced false claims and give the public the right amount of context around it, so that they can hear the claim, but also understand the context for the claim.
These fact check labels are one piece of our approach. But I want to pull back a little bit. My focus is on-- are the security challenges we face. And in particular we found-- you're asking what we can do differently-- it's really important to focus on the threat actors-- the actors who are trying to target public debate and manipulate it.
This type of content-based approach is one piece of the puzzle and one piece of the strategy. But if you only focus on content, that's not going to be enough by itself. Another thing--
- Nathanial, who are the threat actors? Who are they? Are they Russia? Are they China? Are they within our borders?
NATHANIEL GLEICHER: That's a great question. And I would say that there are foreign actors and domestic actors. So whenever we find a network of deceptive behavior that's targeting public debate, we announce it publicly, we share information about it with third party researchers, and we provide details on it.
- Are they not certain governments, though? Can you identify them? It's Russia? It's China?
NATHANIEL GLEICHER: Absolutely. No, absolutely. So we've announced all of these publicly. We've announced so far this year networks from Russia, networks from Iran. We've announced a number of domestic networks, including from political actors right here in the United States and many other government rooms around the world. You can find all of these on our newsroom, where we detail what we've seen and the analysis of what they've tried to do.
The tactics these actors are using is shifting. So what Russia did in 2016 was use large networks of very prominent fake accounts to try to push fake narratives. That doesn't work as well today, because the teams that are trying to find them and expose them at Facebook, and Google, at Twitter, in government, and in civil society, are much better at finding those networks. So we're finding particularly foreign government actors. Instead, they try to use much smaller, much subtler networks that are harder to find.
Now what's interesting here-- a good example of this is an operation called Operation Secondary Infection, which our teams exposed late last year. It was a Russian-backed operation, the most-- the best hidden network of its kind that I've ever seen. But they were still found. They were still caught and exposed. And they didn't get much attention.
And the reason is if you're running a small network that's hidden so it won't get caught, you're running a small network that doesn't get that much attention. And the goal of these networks is to get attention. So the intent here is to force these actors to make a really hard choice. They can run a big large network that'll get caught quickly or they can run a small network that will still get caught, and it won't get that much attention.
That kind of trade off is a difficult place for them to be. They're going to keep trying and we've seen them keep trying. But pushing them in that direction is a good sign and that's the type of goal that all the teams working on this are focused on.
- And Nathaniel, I think the fear is when it comes to cyber security, that the next phase of what they're doing is something you don't see yet, that you're not detecting yet. And so what might that look like? And from a consumer-- I mean, I guess, from a consumer perspective, you don't really know, but from a Facebook security perspective, one would imagine sort of a straightforward hack is almost less of a concern than these more insidious methods that you're talking about. So what's the next generation then? If it's not this small versus big dichotomy, surely, they must be getting more creative?
NATHANIEL GLEICHER: So one of the things we've seen them do is start directly targeting reporters. So as it's harder for them to amplify narratives themselves, they try to trick reporters into doing their amplification for them. We saw a set of Iranian actors use false identities to get op-eds and letters to the editor into a string of real newspapers in the United States.
And we've seen Russian actors directly target reporters, sort of pushing false stories, in hopes that they'll get covered. This happened around the 2018 midterms. And there's a good story there-- the reporters that they reached out to were too smart to get tricked. And the teams that were working to protect the election were able to give enough context and provide enough context about what was happening that the reporter did write a story and the story was Russian actors are pushing false narratives.
But your broader point--
- Make of the broader point, then I have a question about what you're doing to help us all find our polling places.
NATHANIEL GLEICHER: Sure. The broader point that I think is important here is for all the threats that we're prepared for, there will be many threats we haven't thought of yet. And we know that these threat actors will come up with new techniques. So we run red team exercises inside Facebook and with our partners in industry. We actually just a couple of days ago met with government partners and tech companies to talk about the types of threats we're seeing. And all the tech companies have rapid response teams set up, so when the thing happens we haven't thought of yet, we can pivot very quickly to deal with it, working with our government partners, and working with our civil society partners.
- Lastly, part of this new initiative from Facebook-- I know where I usually vote, but now we have early voting in New York state, what you're offering-- will this help me find where I go to early vote? Because it's not the same place.
NATHANIEL GLEICHER: Yeah, so the Voting Information Center is a key part of the broader strategy here. We're running the largest voter information campaign in American history. Our goal is to get 4 million people registered to vote. The voting information centers, which are going to show up in on Facebook and on Instagram will provide accurate and current information about where to vote, how to vote, whether that's early voting or a day of voting.
Also, information about how to register to vote, how to confirm that you're registered. Information about vote by mail, how to request an absentee ballot. And to provide this information right at the top of your news feed and links to posts that talk about voting so that people can find this information quickly and accurately. That's the goal of the initiative.
