AI-generated election misinformation is spreading across social media ahead of the US midterm elections, and experts warn that much of it may be escaping detection as platforms scale back human moderation and rely more heavily on automated systems.
AI tools make it cheap to produce convincing fake images, videos, and claims in large volumes. Meanwhile, major social networks have reduced some of the teams responsible for checking content, leaving automated filters and users to catch more of the falsehoods.
“You can run an information operation now with quality that was impossible in 2016 for super, super cheap,” Frances Haugen, the former Facebook data engineer who became a whistleblower, told The Guardian.
Haugen also explained how people spreading misinformation can evade moderation. By making small changes to each version of a false claim, they can produce posts that get past automated filters even when similar content has already been flagged.
A rumour involving Texas Senate candidate James Talarico shows how quickly such claims can spread.
Earlier this month, posts on X claimed that the Democratic candidate was receiving treatment at a mental health centre in Ohio, when in reality, he was at home with the flu.
The hospital and Talarico denied the claim within hours, La Voce di New York reported. But the rumour continued circulating for days.
Researchers at the Digital Democracy Institute of the Americas, which monitors misinformation targeting Latino voters, found that the story changed as it spread through Spanish-language media.
It evolved into a false claim that Talarico had attempted to take his own life. Correcting the original allegation did not stop the altered version from circulating, researchers said.
Roberta Braga, the institute’s founder, said platforms apply their rules less consistently to Spanish-language content than to English-language posts.
“Now that they’re using AI to do a lot of the content moderation, even more is likely slipping through the cracks,” she said.
AI-generated content is also appearing in official campaign advertising. Ken Paxton, the Republican challenging Talarico, ran an advertisement featuring AI-generated images of the Democratic candidate.
The ad disclosed its use of AI, and some states have introduced rules requiring such disclosures, including Pennsylvania.
The major platforms have different approaches to political misinformation, but their rules cover only certain types of content. Meta labels AI-generated images it can identify and requires political advertisers to disclose AI use.
YouTube prohibits false claims about when, where, and how people can vote. X relies largely on its crowdsourced Community Notes system.
Many other misleading political claims fall outside these specific rules.
Meta and X did not respond to The Guardian’s requests for comment. YouTube said the same guidelines apply to all users, including political candidates.
The issue extends beyond the US. In Spain, a political party is facing a court complaint over a deepfake election video.
As election campaigns increasingly use AI to create and distribute content, the challenge for platforms is not simply identifying synthetic media. It is also catching false claims that change as they spread, across languages and communities, faster than corrections can follow.