US AI politics split
U.S. lawmakers across parties say they want more AI regulation but can’t agree on what the problem is or what to regulate. (reason.com). The piece argues regulation is likely to show up through adjacent channels—consumer protection, platform liability, election law and procurement—rather than a single federal AI statute. (reason.com)
Congress keeps talking about regulating artificial intelligence, but lawmakers are still fighting over what exactly needs policing. (reason.com) A April 11 Reason review of federal bills found a party split in approach: many Democratic proposals target harms such as deepfakes, fraud and election abuse, while several Republican proposals focus more on the underlying models, federal preemption and limits on state rules. (reason.com, whitehouse.gov) Congress has moved faster on narrow bills than on a single artificial intelligence law. The TAKE IT DOWN Act, Senate Bill 146, became Public Law 119-12 on May 19, 2025, requiring covered platforms to remove nonconsensual intimate images, including computer-generated ones, after notice. (congress.gov, govtrack.us) That pattern has held across other fights. The NO FAKES Act of 2025, introduced in the Senate on April 9, 2025, would create rights over a person’s voice and visual likeness and set up a notice-and-takedown system for unauthorized replicas. (congress.gov, legiscan.com) The federal backdrop also changed when President Donald Trump revoked President Joe Biden’s 2023 artificial intelligence executive order on January 20, 2025. A March 20, 2026 White House framework then called for a national policy that protects rights, supports innovation and avoids a “patchwork” of state laws. (federalregister.gov, govinfo.gov, whitehouse.gov) House leaders had already pointed in a piecemeal direction. The bipartisan House Artificial Intelligence Task Force, created on February 20, 2024 and reporting on December 17, 2024, said Congress should favor sector-specific rules and incremental updates over one sweeping statute. (speaker.gov, science.house.gov) That leaves existing agencies with a large role. The Federal Trade Commission has said there is no artificial intelligence exemption from consumer protection law, and it has kept using unfair-and-deceptive-practices authority to scrutinize artificial intelligence claims and harms. (ftc.gov, data.aclum.org) Public opinion is less divided than Congress on the basic question of whether government should act. Ipsos said in March 2026 that 63 percent of Americans want the federal government to ensure artificial intelligence outputs do not cause harm, including 67 percent of Democrats and 63 percent of Republicans. (ipsos.com) So the next federal artificial intelligence rules are more likely to arrive disguised as election law, platform liability, consumer protection and federal purchasing standards than as one landmark code. Congress agrees there is a problem; it still does not agree on which problem comes first. (reason.com, speaker.gov, whitehouse.gov)
