Artificial intelligence-generated deepfakes have become a significant threat to democratic elections and public discourse, as demonstrated by a viral fake image of a Georgia mayor that spread on Facebook before being debunked. The ease with which AI image generators can now create convincing fabricated imagery—tasks that previously required advanced technical skills—has alarmed legislators, advocates and the public alike, with 71% of Americans expressing concern that AI-generated deepfakes could affect elections. Beyond electoral threats, concerns about deepfakes being used to sexualize women and children outweigh worries about AI's economic impact by roughly three to one.

The Trump administration has already deployed altered images, including a doctored photograph of activist Nekima Levy Armstrong in which her skin was darkened and she was falsely depicted as crying. Senate Democrats are developing contingency plans to address potential election interference scenarios involving fabricated images and videos circulating on social media, whilst Senator John Kennedy has recently proposed legislation requiring artificial intelligence firms to install "kill switches" for rogue AI systems. Senator Raphael Warnock encapsulated the democratic concern: with hours before an election, fabricated images and false stories could spread widely, leaving voters unable to distinguish fact from fiction and fundamentally undermining electoral integrity.