Donald Trump is set to return to the White House in his most successful presidential election to date. But in an election with such a significant online presence, what role has misinformation and disinformation involving artificial intelligence had to play in its outcome?
Misinformation and disinformation are not exclusive to the 2024 election; it’s widely accepted that Russia interfered back in 2016, by hacking Hillary Clinton’s emails and spreading propaganda on social media. The Senate Select Committee on Intelligence found that the Internet Research Agency (IRA), described as a “troll farm”, had posed as “US political activists” spreading misinformation on social media, which was chiefly used to the “detriment of Secretary Clinton’s campaign.” This included the sharing of hashtags, such as “Hillary4prison”. In 2019, Robert Mueller’s report on Russian interference found no evidence to suggest that the Trump campaign had been involved.
The 2024 election saw the use of new examples of misinformation and disinformation; taking the form of artificial intelligence. Much of this AI misinformation has come from Russia, but not entirely. Some US citizens have also been responsible for the spread of AI misinformation. In February 2024, the Federal Communications Commission (FCC) made it illegal to generate AI voices in “robocalls”, after voters in New Hampshire were discouraged from voting in the primary by a deepfake of President Biden. The call was actually commissioned by a Democratic political consultant, supposedly taking just twenty minutes to create.
But deepfakes are not just audio. Deepfake images of Donald Trump have also been shared this election, primarily by Trump and his supporters. Unlike audio, there are more discrepancies found in images. Earlier this year, two images of Trump went viral, and were very quickly recognised as being AI-generated; given away by the shiny skin and inconsistencies with people’s fingers. While these images, and others shared by Trump, are clearly fake, what they represent still promotes misinformation. Trump has regularly shared deepfakes of himself, including one which showed him riding a lion. No one is suggesting that this particular image is real, but it is being used as a propaganda tool to portray Trump in a certain way.
Pew Research Center found that 57% of US adults were “extremely/very” concerned about the influence of AI on this election. The Institute for Strategic Dialogue highlighted that this concern is valid; finding that X (previously known as Twitter) users were “misidentifying content in 52% of cases”, unable to distinguish between what was real and what was fake. Misinformation, including AI-generated content, may not have decided this election, but it certainly had a role to play. Given that the party most responsible for the sharing of AI-generated content now has control of every branch of the US federal government, it is certainly not going away any time soon.