Electoral Integrity & Political AI

Elections are a particularly high-stakes arena for AI. The technology enables several new threats to electoral integrity: AI-generated deepfakes of candidates saying things they never said, synthetic robocalls impersonating officials, automated social media accounts spreading disinformation at scale, and micro-targeted messaging designed to suppress voter turnout in specific communities. These aren't hypothetical - examples have appeared in elections worldwide. The challenge for election authorities is that AI-generated content is becoming increasingly difficult to distinguish from authentic material, and it can be produced and distributed faster than fact-checkers can respond. Some jurisdictions are introducing specific rules around AI in elections: mandatory disclosure of AI-generated political advertising, bans on deepfakes of candidates in the period before elections, and requirements for platforms to label synthetic content. Watermarking and provenance technologies like C2PA offer technical solutions, but adoption is still early. For organisations involved in political communication, advertising, or media, the rules around AI and elections are tightening quickly. Even where specific legislation hasn't passed, deploying AI in ways that undermine electoral integrity carries enormous reputational and legal risk.