Cheap AI-generated 'deepfakes' are transforming US political attack ads ahead of the November midterms. Texas Democratic Senate nominee James Talarico has been depicted in fabricated videos recanting real comments, reading old social media posts, and even appearing as a masked doctor. His Republican opponent Ken Paxton is behind one such ad.
The Wesleyan Media Project identified 164 political ads—"almost certainly an undercount"—generated or enhanced with AI, with roughly $80mn spent on them. These range from satire to convincing deepfakes where candidates say or do things they never did. More than 30 states have passed laws restricting AI in political ads or requiring disclosure, but many face constitutional challenges.
Nearly 70% of the AI ads found lacked disclaimers. Campaigns are using the technology to produce ads that would have cost six figures to create traditionally.
Source: Financial Times Companies · Summarized by HeadlinesBriefing