FAIR Elections Act (S.4774): regulating election-deceiving AI (deepfakes and the like)

1Elections depend on accurate information

Elections, the foundation of democracy, presume that voters can decide based on accurate information. But generative AI can craft fake video and audio (deepfakes) putting words a candidate never said into their mouth, so cleverly they are hard to tell from the real thing. If such disinformation spreads right before a vote, it can deceive voters and distort the outcome. This is a new threat AI poses to democracy, watched warily in many countries.

2Addressing fraudulent AI content in elections

S.4774, the FAIR Elections Act (formally the Fraudulent Artificial Intelligence Regulations Elections Act), is a Senate bill that seeks to address this threat. As the acronym FAIR (Fraudulent Artificial Intelligence Regulations) indicates, it appears aimed at regulating the use of AI to deceive voters in elections.

The sponsor is Sen. Jeff Merkley (D-OR); it was introduced on June 11, 2026, and referred to the Senate Committee on Rules and Administration, which oversees elections. What AI uses would be regulated and what disclosure requirements or bans would fall on whom are provision details beyond the information here, so no firm conclusions are drawn.

3Correcting afterwards does not arrive in time

An election presupposes that voters can judge on accurate information. Because disinformation spreads fast and wide, where the response is placed changes what it means.

Generative AI can produce fake video and audio putting words a real candidate never said into their mouth, indistinguishable from the real thing. The acronym FAIR stands for Fraudulent Artificial Intelligence Regulations. The bill was introduced on 11 June 2026 and referred to the Senate Rules Committee, which owns elections.