AI-Generated Deepfakes Threaten U.S. Political Reality

In the May 19 Kentucky Republican primary, the super PAC “MAGA Kentucky” released an AI-generated video falsely alleging Rep. Thomas Massie was in a romantic “throuple” with Ilhan Omar and Alexandria Ocasio-Cortez. The ad said, “Thomas Massie, caught in a throuple in Washington.” Massie called the clip a desperate lie aimed at older voters. He lost to Trump-endorsed challenger Ed Gallrein, a defeat analysts note the AI attack may have contributed to.

Salon

AI creates “slopaganda,” a term from philosophers Mark Alfano and Michal Klincewicz for disinformation. RNC released a video showing dystopian outcomes after a Biden win; Trump has shared clips. Brookings fellow Darrell West warned AI “democratizes disinformation” and equips “troll farms.”

Salon

Key participants are Rep. Thomas Massie, Kentucky incumbent; the “MAGA Kentucky” super PAC that created the deepfake; Democrats Ilhan Omar and Alexandria Ocasio-Cortez; Brookings fellow Darrell West, who warned of AI-enabled “troll farms”; and Steven Rosenbaum, who critiques bans. Legislatures in 31 states have passed deepfake rules.

Salon

A 2024 Pew Research poll found 82 % of Americans fear AI-generated political misinformation. By 2026, 31 states have enacted deepfake restrictions; Texas imposes a 30-day pre-election ban, while California, Maryland and Massachusetts require clear AI-disclosure labels on political ads.

Salon

Massie called the video a “desperate lie.” West warned AI “democratizes disinformation” and threatens. Rosenbaum said bans are “well-intentioned, but ultimately they fail” as tech firms outpace law. Congressional Democrats have introduced disclosure bills, none passed. Trump postponed an executive order on AI national-security risks, citing competitiveness concerns.

Salon

Civil-rights groups warn that AI-driven “hypernormalization” erodes public trust and fuels authoritarian appeal. Libertarian voices argue that broad bans infringe free-speech rights. Critics note many state bans apply only in narrow pre-election windows, limiting deterrence.

Salon

Analysts differ on state-law impact: some view disclosure mandates as a success in raising awareness, others deem them largely ineffective without enforcement. No comprehensive database tracks the volume of AI-generated political ads, leaving the scale of the problem uncertain.

Salon

• “It reeks of desperation, but they’re hoping the older generation won’t realize it’s an AI generated lie.” — Thomas Massie, Rep. (KY)

Salon

• “Thomas Massie, caught in a throuple in Washington,” — Narrator, AI-generated ad

Salon

• “democratiz[ing] disinformation and giv[ing] everyone the capabilities of troll farms.” — Darrell West, Brookings senior fellow

Salon

• “well-intentioned, but ultimately they fail. The problem is these technology companies are incredibly agile . . . I don’t think we’re going to solve the problem by legislating away particular apps.” — Steven Rosenbaum, author

Salon

The 2026 midterms will likely see more AI-crafted attacks as deepfake tools become cheaper. Lawmakers are expected to revisit disclosure legislation, and the White House may revive the postponed AI-security executive order. Election-security agencies plan to expand monitoring of AI-generated political content.

Salon