Trust-and-safety teams have lost staff while platforms increasingly use AI for moderation
Social media feeds are awash with what experts cited by The Guardian describe as an unprecedented flood of AI-driven disinformation ahead of November’s US midterm elections, with many posts likely slipping through the cracks. Tens of millions of Americans are preparing to cast their votes, and social media companies have scaled back efforts to protect voters from lies meant to manipulate elections. The Guardian identified Meta among the companies that have pulled back on that voter protection. People cast their votes on the first day of early voting in New York City on Saturday, in a photograph published with The Guardian’s reporting (Heather Khalifa/AP).
The Guardian reported that the backlash against Facebook after the Cambridge Analytica scandal prompted social media companies to bolster their trust and safety operations, and that a political backlash over fears of speech restrictions, which The Guardian said intensified after Elon Musk’s takeover of X, formerly Twitter, and the election of Donald Trump in 2024, has led to significant layoffs from those teams. Separately, the same reporting said artificial intelligence has made it easier than ever to launch a disinformation campaign and that platforms are increasingly relying on AI, rather than human intervention, to respond.
The Guardian reported that organized groups behind election disinformation once operated on the fringes and that their work might now be amplified by official government sources. Viral posts on X and rightwing blogs claimed that an undocumented immigrant arrested in central Pennsylvania in an alleged public indecent exposure case was an active field worker and canvasser for Democrat Janelle Stelson’s congressional campaign. As campaign staffers set the truth sprinting to catch up with the lie, an official statement from the Department of Homeland Security amplified the claim. “It’s unbelievable that Janelle Stelson would employ an illegal alien to work on a congressional campaign, let alone a pervert who exposed himself in public,” declared Charles Wall, the deputy director of Immigration and Customs Enforcement (ICE).
The campaign released records showing that the immigrant, Chimaobim Adedayo Igbojekwe, had no paid position on the campaign, despite an initial police incident report stating that Igbojekwe claimed to be a campaign worker. The Stelson campaign holds the claim provably false, and the campaign of Stelson’s opponent, Republican congressman Scott Perry, continues to run ads with the claim. The Stelson campaign sent cease-and-desist letters to the media, and a Fox News reporter took down a post on X about the indecent exposure arrest. Activists at Citizens for Responsibility and Ethics in Washington launched a complaint accusing DHS of violating the Hatch Act, which limits the political activities of government employees. “Trump has made clear that he intends to interfere in the elections,” said Nora Benavidez, a civil rights and free speech attorney with Free Press, a nonprofit advocating for media independence. Trump “has leveraged agencies across the federal government, whether it’s the US Postal Service or DHS – you name it,” Benavidez said. The Stelson campaign has yet to complain directly to the social media platforms, including Meta’s Facebook, Elon Musk’s X and Google’s YouTube, though a campaign spokesperson said the campaign “continues to consider every legal avenue to fight the false claims in Perry’s ads.” A photograph accompanying the reporting shows Stelson at a campaign event in York Haven, Pennsylvania, on 23 September (Marc Levy/AP).
James Talarico, the Democrat running for US Senate in Texas, took time off from campaigning earlier this month to recover from the flu. Many social media feeds supplied another, false, reason for his absence: that he was being treated at a mental health facility in Ohio. The Guardian reported the claim was a complete fabrication, pushed by a rightwing “provocateur” who published a deceptively edited phone call with a hospital staffer. Talarico’s campaign released security-camera video showing him at home in Texas, and the candidate publicly clarified that he had never even been to Ohio. The lie circulated, and was stoked, for several days. A photograph accompanying the reporting shows Talarico speaking at the Voto Latino ‘Ready, Go, Vote,’ rally in Weslaco, Texas, on Sunday before the midterm elections (Michael Gonzalez/AP).
The Digital Democracy Institute of the Americas tracked how the claim propagated and how the misinformation mutated as conservative Spanish-language media picked it up. What began as a viral lie about Talarico being hospitalized in Ohio became a false claim in Spanish that he had attempted suicide. DDIA’s researchers wrote that the case “illustrates a recurring challenge in countering election disinformation.” Lies evolve and shift as they spread, they wrote, “making the original correction insufficient to address every version of the story.”
Roberta Braga, the founder and executive director of DDIA, said social media platforms enforce their terms of service “more congruently” in English than in other languages including Spanish. “There’s a lot less guardrails in place from the companies to deal with Spanish language, especially now that they’ve walked back a lot of their own investments in content moderation, and they’ve moved toward doing community-centered content moderation, like Community Notes-style things,” Braga said. “Now that they’re using AI to do a lot of the content moderation, even more is likely slipping through the cracks.”
Frances Haugen, a former Facebook data engineer turned whistleblower and advocate for transparency and accountability in Big Tech, said AI has made it easier than ever to launch a disinformation campaign. “You can run an information operation now with quality that was impossible in 2016 for super, super cheap,” she said. Operatives playing dirty tricks can get around content moderation rules, she said, “because you can just keep tweaking the post a little bit each time you create them, and some of them are going to get through.”
When content can be tweaked until it gets through, Haugen said, the focus has to shift to distribution. “When you stop being able to focus on content because the AI will just slightly tweak it until it gets through, you have to focus on what are the systems of distribution,” she said. “What are the biases? What are the ways we have feedback systems, cycles, that allow the community to provide feedback, that allows it to be damped down?” Campaigns from Pennsylvania to Texas have been scrambling to respond to lies that have surfaced and spread rapidly online, and unlike previous elections they find themselves in a quandary, The Guardian reported: who, exactly, do they turn to when fabrications and falsehoods propagate?
Ken Paxton, Talarico’s Republican opponent in Texas, launched a campaign advert earlier this year using AI-generated footage of Talarico, a Presbyterian seminarian, supposedly rewriting the Bible. “I think Texans are fed up with these kind of lies in our politics,” Talarico told NBC News in September. “These AI deepfake ads don’t add anything to our civic discourse. They don’t do anything to elucidate the choice that’s in front of voters.”
The dominant social media platforms do have clear rules, based on specific and limited definitions, for handling misinformation driven by AI during election campaigns. Meta requires advertisers to disclose the use of synthetic images and AI-generated content with election ads, and enforces automated labeling on AI-generated images identified through Coalition for Content Provenance and Authenticity (C2PA) metadata. Meta bans posts and ads that misrepresent voting dates, locations or legal requirements, as well as threats of violence. Beyond excluding unsolicited political content from its recommended posts, the company does little active moderation of political commentary. An analysis by Free Press noted that Meta narrowed its historical restrictions in 2023 to allow debate around the legitimacy of past election cycles.
In 2023, YouTube stopped removing content that falsely claimed the 2020 election, or other past US presidential elections, were marred by “widespread fraud, errors or glitches,” saying the removals had curbed misinformation but also had a chilling effect on political speech. The platform defines election misinformation narrowly: content that misleads voters about the time, place, means or eligibility requirements for voting, and false claims that could materially discourage voting, are banned, as are claims that a candidate on the ballot is ineligible, calls to hack a polling site’s computers and calls to incite violence at election locations. “All users” of YouTube, including political candidates and campaigns, “are subject to the same community guidelines … and have access to the same public reporting tools,” according to the platform. YouTube’s spokesperson wrote in response to inquiries that “We also review government requests closely to determine if content should be removed because it violates a law or our policies.”
Google operates a likeness detection system and allows civic leaders to request removal of a video when an AI-generated image falsely depicts their words or deeds, but a direct lie that falls outside those lines could propagate without intervention, The Guardian reported. Instead of relying on takedowns, Google prefers to amplify authoritative news outlets and election boards in search queries and underneath high-stakes civic videos through contextual “information panels.”
X, formerly known as Twitter, abandoned internal safety reviews after its acquisition by Musk, replacing them with the crowdsourced “community notes” program. Its civic integrity policy bars false information about how to vote and suppression schemes. The platform also supposedly restricts synthetic media designed to mislead, though experts argue that the policy amounts to little without enforcement.
Meta Platforms, the Facebook and Instagram owner, and X did not respond to requests for comment. YouTube, owned by Google, said all its users are subject to the same guidelines. The Guardian’s reporting leaves the campaigns it describes in a quandary: who, exactly, do they turn to when fabrications and falsehoods propagate?