A fake video of House Speaker Nancy Pelosi (D-Calif.) posted to Facebook on Thursday that was edited to make her appear drunk is underscoring a quickly evolving danger for 2020 campaigns.
One cycle after Russia’s interference in the election through hacks and fake posts on social media wreaked havoc, presidential candidates now have to worry about videos doctored by artificial intelligence technologies that can make candidates say things they didn’t say or look completely different.
The video of Pelosi posted to Facebook didn’t use such advanced technologies.
{mosads}It was slowed down to make Pelosi appear unwell or potentially drunk, and it still succeeded in fooling many people — at least judged by comment boards.
The video had been viewed more than 2.5 million times as of Friday afternoon. Facebook is refusing to remove the clip, saying it doesn’t violate platform guidelines, though it is not recommending the video in its news feed.
President Trump waded into the controversy, tweeting a clip from Fox Business Network that compiled every time Pelosi stumbled over her words during a recent press conference. The clip included a Fox News commentator saying Pelosi appeared “worn down.”
Coupled with remarks by Trump, it at least appeared the president was trying to use the fake video to his advantage to build a narrative that the Speaker “had lost it,” in Trump’s words. But Trump denied knowing about the fake video.
The fake Pelosi video points to a danger that experts have been offering increasing warnings about.
Fabrice Pothier, a senior advisor for the Transatlantic Commission on Election Integrity, said foreign actors could use fake videos “to sow distrust and decredibilise their opponents.”
Pothier said the Pelosi video is essentially a cheap one, and more advanced fakes are coming.
“We are still at a rudimentary stage — the Pelosi video is a basic alteration of an authentic video — but at this pace, it is only a matter of time before fully synthetic video and audio files — deepfakes — generated by algorithm rather than with video editing tools contaminate our information sphere,” Pothier said.
John Villasenor, a senior fellow at the Brookings Institution, said the use of such videos in campaigns is “inevitable.”
“Unfortunately, deepfakes are going to be a part of the political landscape in the 2020 campaign and beyond,” Villasenor said. “It’s an inevitable next step in the information manipulation that we saw in the 2016 campaign.”
It’s unclear if the Pelosi video will have that much of an impact.
But Charlotte Stanton, the director of the Carnegie Endowment for International Peace’s Silicon Valley office, warned a similar video — timed just right — could have a calamitous effect on an election.
“Can you imagine if an altered video like this of a presidential candidate surfaced a day before the election?” she told The Hill.
She said many campaigns would not be ready to fight back at a fake video and that even if they were, it could quickly become too late.
“By the time the campaign has caught a fake, the fake will have circulated across social media,” she said, arguing it was up to Facebook and other social media platforms to better police their content.
YouTube did say it would take down any altered videos involving Speaker Pelosi and that the fake video accumulating views violated its policies.
Facebook said it would flag the Pelosi video as potentially fake news but would not remove it.
“In this particular case, this video would be eligible for fact-checking from one of our third-party fact-checking partners,” a Facebook spokesperson told The Hill on Friday. “We did enqueue this to fact-checkers for review and as of yesterday evening, one of our fact-checking partners reviewed the video and rated it ‘False,’ so we are now heavily reducing its distribution in News Feed and showing additional context from this fact-checker in the form of a “Related Articles” unit in News Feed where it still appears.”
Sen. Brian Schatz (D-Hawaii) criticized Facebook’s response, tweeting on Friday that “Facebook is very responsive to my office when I want to talk about federal legislation and suddenly get marbles in their mouths when we ask them about dealing with a fake video. It’s not that they cannot solve this; it’s that they refuse to do what is necessary.”
Deepfake videos have caught the attention of lawmakers at both the federal and state level. This week, the Texas House of Representatives passed legislation that would criminalize creating a deepfake video with the intent to “injure a candidate or influence the result of an election.” The Texas Senate already passed this bill in April, and it now awaits the governor’s signature.
At the federal level, Sen. Ben Sasse (R-Neb.) introduced a bill at the end of the last Congress that would penalize a person who created or distributed “fraudulent audiovisual records.” A spokesperson for Sasse did not respond to request for comment when asked if he planned to reintroduce the bill this Congress.