State AI Laws in 2026: Where the Real Rules for Superintelligence Live
Washington has chosen voluntary pledges over new federal rules. That leaves the states as the place where binding AI law is actually being written.
No federal AI law, yet
The U.S. still has no comprehensive federal law covering AI. In July 2025 the Senate voted 99 to 1 to strip a proposed ten-year freeze on state AI enforcement out of a major budget bill. A December 2025 executive order told the Justice Department to challenge state AI laws, but an order cannot cancel them on its own, and as of spring 2026 no court had struck one down.
The key state laws
- California, SB 53 (frontier AI transparency). In effect since January 1, 2026. The biggest frontier model developers must publish a safety framework and report serious incidents, with fines of up to $1 million per violation.
- Texas, TRAIGA. Also in effect since January 1, 2026. It bans certain harmful uses of AI and is enforced by the state attorney general.
- Illinois, HB 3773. Since January 1, 2026, it covers AI used in hiring and other job decisions, and lets individuals sue.
- Colorado AI Act. Covers high-risk AI in major life decisions like jobs, housing and loans. After a delay it was scheduled to start June 30, 2026.
- New York, RAISE Act. A frontier model safety law aimed at 2027.
The bigger picture
Lawmakers filed roughly 1,200 AI bills in 2025, and about 145 became law across 38 states, according to trackers cited by the Cloud Security Alliance. Expect more in 2027 sessions, especially if the federal Super Intelligence Force's report sticks with the "no new law" approach.