Those fears appear to have been unwarranted, says Sam Stockwell, the researcher on the Alan Turing Institute who conducted the study. He focused on three elections over a four-month period from May to August 2024, collecting data on public reports and news articles on AI misuse. Stockwell identified 16 cases of AI-enabled falsehoods or deepfakes that went viral through the UK general election and only 11 cases within the EU and French elections combined, none of which appeared to definitively sway the outcomes. The fake AI content was created by each domestic actors and groups linked to hostile countries reminiscent of Russia.
These findings are in keeping with recent warnings from experts that the deal with election interference is distracting us from deeper and longer-lasting threats to democracy.
AI-generated content seems to have been ineffective as a disinformation tool in most European elections this 12 months up to now. This, Stockwell says, is because most people who were exposed to the disinformation already believed its underlying message (for instance, that levels of immigration to their country are too high). Stockwell’s evaluation showed that individuals who were actively engaging with these deepfake messages by resharing and amplifying them had some affiliation or previously expressed views that aligned with the content. So the fabric was more more likely to strengthen preexisting views than to influence undecided voters.
Tried-and-tested election interference tactics, reminiscent of flooding comment sections with bots and exploiting influencers to spread falsehoods, remained far more practical. Bad actors mostly used generative AI to rewrite news articles with their very own spin or to create more online content for disinformation purposes.
“AI is just not really providing much of a bonus for now, as existing, simpler methods of making false or misleading information proceed to be prevalent,” says Felix Simon, a researcher on the Reuters Institute for Journalism, who was not involved within the research.
Nevertheless, it’s hard to attract firm conclusions about AI’s impact upon elections at this stage, says Samuel Woolley, a disinformation expert on the University of Pittsburgh. That’s partially because we don’t have enough data.
“There are less obvious, less trackable, downstream impacts related to uses of those tools that alter civic engagement,” he adds.
Stockwell agrees: Early evidence from these elections suggests that AI-generated content could possibly be more practical for harassing politicians and sowing confusion than changing people’s opinions on a big scale.
