First came the election results, then came the meme party.
In the wake of the Democratic wave of support on Election Day, content creators used artificial intelligence to flood the internet with fake videos, depicting candidates and celebrities in various states of celebration.
One A.I.-made skit, featuring replicas of the cast from the hit TV show “The Office” showed the boss adjusting to a progressive society, wearing a Pride flag sweater and calling for an emergency Diversity Day. It received more than 16 million views in one post on X.
The video was emblematic of the kinds of A.I. fakes that have coursed through social media during the midterm season. Many of the most popular videos spreading online vied for laughs and sometimes jeers — an absurd mix of outlandish events that obviously never happened.
The results are a minor relief for researchers who monitor the spread of A.I. fakery, and also a reckoning.
Experts have warned ahead of every U.S. election since at least 2020 that A.I. replicas of real people — “deepfakes” — and other deceptive content generated by the technology could upend democracy, misleading voters about scandals that never happened or salacious fake comments.
Some ads published this year depicted realistic A.I. versions of candidates. But far more of the videos circulating online were parodies, jokes and memes that were too absurd to be believed.
“The case that everyone’s worried about is a highly inflammatory or incriminating deepfake going viral on the eve of an election and truly moving the needle on how people have voted,” said Henry Ajder, an expert on A.I. and the founder of Latent Space Advisory, an A.I.-consulting firm. “That has not yet come to bear.”
Memes have spread widely online showing real candidates in bizarre situations, like one series mocking President Trump as a small business owner with Vice President JD Vance as his son and employee. Creators who opposed Senator-elect Abdul El-Sayed’s bid released a flurry of videos disparaging the candidate, often deploying racist stereotypes.
Hundreds of A.I.-generated campaign ads also flooded the internet and airwaves, including some that were deceptive and not disclosed as A.I.-made, according to a report by the Wesleyan Media Project, a research group at Wesleyan University. A vast majority of the ads were obviously fake, depicting impossible scenes.
Some were more grounded in reality, including an ad showing an A.I. version of Vikki Goodwin, then a Democratic candidate for lieutenant governor in Texas, purportedly supporting a “death tax.”
“There are a lot of people who don’t know me yet, who might think this is me,” she told The New York Times in September. The campaign behind the ad said at the time that it was “an obvious parody produced to entertain.”
Ms. Goodwin ultimately won her race by more than six percentage points.
One frequent target of attack ads, Senator-elect James Talarico, a Texas Democrat, won his election handily. Though there is plenty of material available of Mr. Talarico with which to generate A.I. fakes or audio clones, experts could not find evidence of a deceptive deepfake of him that posed as genuine footage. Instead, the candidate was featured in dozens of A.I.-generated videos dancing and singing, which were viewed millions of times online.
Those A.I. fakes can still have political power, rallying supporters and bringing to life criticisms made against candidates, which can stick in voters’ minds.
“That’s something that political campaigns have done forever — that’s nothing new,” Mr. Ajder said. “The means of production are new, the ability to scale, the access — the viscerality is different.”
The memes match content published during primary elections in the United States and in recent general elections in other countries. A report from Canadian researchers at McGill University and the University of Montreal found that A.I.-generated content made up about 6 percent of election-related images on X during the country’s federal election last year. But more than 90 percent of the views went to more innocuous content, like memes, jokes and fake depictions of candidates as superheros or supervillains.
Experts have pointed to a handful of past cases where A.I.-generated fakes seemed to mislead voters. In 2023, an A.I.-generated audio clip was credited with potentially turning the course of an election in Slovakia. In 2024, an A.I. replica of President Joseph R. Biden’s voice was used to call voters in New Hampshire, advising them to skip the primary election.
This year, a political group created a phone line for an A.I.-powered version of Nithya Raman, the Los Angeles mayoral candidate who ultimately won her race. On the phone line, she had appeared to endorse tent encampments.
Such fakes tend to target partisans who are more likely to believe them, rather than reaching swing voters who tend to show more skepticism. They may be spread more easily in group chats or private groups, researchers said.
And online videos with a more comedic edge have collected millions of views.
“The threat of deepfakes has not disappeared,” said Shaina Raza, a research scientist at the Vector Institute of Artificial Intelligence, who has studied A.I. risks. “It is taking different forms than what we expected.”
