Politicians are increasingly pushing AI-generated content. It's scaring other politicians. (www.businessinsider.com)

AI-generated images and videos are rapidly becoming a staple in U.S. political communication, with politicians and campaigns using hyperrealistic synthetic media to mock opponents, amplify messages and—critics say—blur the line between parody and deception. Recent examples include an AI-made video reposted by President Trump showing a fabricated scene of him dumping a brown substance from a jet, AI attacks by Andrew Cuomo in the NYC mayoral race, and a Senate GOP clip that stitched a real Chuck Schumer quote into a fabricated video to boost reach from 100k to 1.8M views. Some lawmakers treat clearly labeled parody as acceptable, but several — including Sen. Chris Murphy and Sen. Mark Kelly — warn that indistinguishable deepfakes threaten civic trust and meaningful political dialogue. Technically, modern generative models can produce photo-realistic faces, lip-syncing and context-aware edits at low cost, making synthetic political content both scalable and viral. That raises policy and security dilemmas: how to deter deception without trampling First Amendment protections. Proposed responses range from mandatory, indelible watermarks and narrow bans (e.g., the Protect Elections from Deceptive AI Act) to relying on fraud statutes, while some lawmakers (e.g., Ted Cruz) argue existing free-speech safeguards should limit regulation. The convergence of low-cost generative tools, viral amplification, and adversary misuse makes detection, provenance tracking and platform policy urgent priorities for the AI/ML and public-policy communities.