Deepfakes and AI Memes: Their Impact on Global Elections in 2024

U.S. elections in 2024 saw AI-generated content reshape political discourse, though not in the catastrophic way many experts had feared. While the dreaded “bombshell deepfake” scenario largely failed to materialize, artificial intelligence tools instead found their way into mainstream political communication through memes and manipulated media that were often openly shared by candidates and their supporters.

The year began with a troubling sign when thousands of New Hampshire voters received robocalls featuring what sounded like President Biden telling Democrats not to vote in the upcoming primary. The voice urged listeners to “save your vote for the November election.” This deepfake, later revealed to be commissioned by a Democratic political consultant who claimed he wanted to raise awareness about AI dangers, resulted in a $6 million FCC fine and criminal charges.

“The nightmare situation was the day before, the day of election, the day after election, some bombshell image, some bombshell video or audio would just set the world on fire,” said Hany Farid, a professor at the University of California at Berkeley who studies manipulated media.

Yet as 2024 progressed, the pattern that emerged was less about deception and more about what Farid calls “death by a thousand cuts” – a steady stream of AI-generated content designed to influence rather than deceive outright.

“I don’t think the images were designed to be clearly deceptive, but they were designed to push a narrative, and propaganda works,” Farid explained. “I do think that there was a general polluting of the information ecosystem where people just started to give up.”

This trend extended beyond the United States. In Indonesia, the political party Golkar used AI to digitally resurrect former dictator Suharto, who died in 2008. The AI-generated Suharto endorsed the party’s candidates, saying they would “continue my dream of Indonesia’s progress.” Shortly after, Suharto’s son-in-law, who had Golkar’s backing, won the presidential election.

India’s election, the world’s largest democratic exercise, saw widespread distribution of AI-generated memes via WhatsApp. One Indian trader told NPR he enjoyed memes mocking opposition leader Rahul Gandhi, including one depicting an AI version of Gandhi as a thief imagining money he would steal if elected. However, the trader noted the memes didn’t influence his voting decision.

Sahid SK, an AI artist in India, suggested creators favored memes over deepfakes due to legal concerns: “I think that’s the only reason we haven’t seen many deepfakes in this election. Because everybody’s afraid of legal notices.”

In the U.S. presidential race, AI-generated content became particularly visible through social media platforms. Elon Musk, owner of X (formerly Twitter), shared a fake advertisement featuring an AI clone of Vice President Kamala Harris’s voice describing herself as “the ultimate diversity hire,” without disclosing it was originally posted as parody content.

Musk and other supporters of former President Donald Trump frequently shared AI memes mocking Harris and Democrats while portraying Trump in heroic imagery. Trump himself posted an AI-generated image suggesting Taylor Swift had endorsed him, which was false.

Zeve Sanderson, who studies digital information at NYU’s Center for Social Media and Politics, noted that while this content wasn’t designed to change minds, it served to “make their preferred candidate look patriotic or noble [or] to make their opposing candidate look evil.” Traditional photo and video editing tools could achieve similar results, but “generative AI just makes it quite a bit easier.”

Experts caution that AI may have been used in less detectable ways throughout the election cycle. “I still don’t think we have a really good, rigorous sense of how generative AI was used at scale, by whom, for what purposes,” Sanderson said.

While Farid believes AI-generated content likely didn’t change election outcomes, he acknowledges its impact on public perception: “Do I think it changed the outcome of the election? No. Do I think it impacted people’s thinking? Yeah, I think it did. And I think it will continue to do that.”

As generative AI technology becomes more accessible and sophisticated, the challenge of maintaining information integrity during elections will likely become even more pronounced in future electoral cycles worldwide.