US states introduced 1,539 AI bills in 2026, up from 1,262 a year earlier, according to the National Conference of State Legislatures ( NCSL ).
That volume hasn’t produced a corresponding wave of comprehensive law. By early October, states had enacted 280 of those bills, with the remainder targeting narrower subjects such as deepfakes, notification and frontier-model audits. Only Colorado SB 189 and Connecticut SB 5 qualified as broad frameworks.
For companies operating across state lines, the result is a shifting set of obligations.
Notification legislation, which requires platforms to tell users they’re interacting with AI, saw the sharpest increase. NCSL fellow Will Clark put the rise at nearly three times the previous year’s level. He described the category as “letting a user know that they’re engaging with AI.”
Deepfake bills were the most common type enacted in 2026, typically regulating explicit content, political advertising, and the use of digital likenesses without consent. Healthcare and education also drew attention from state lawmakers.
Clark said state action on AI mirrors many of the actions taken by the federal government, though the volume of legislation is far higher at the state level. Private-sector use of AI topped the list of proposals in 2025, followed by deepfakes and government use. He noted that introduced bills don’t necessarily become law.
Congress has deliberated bills including the Kids Online Safety Act and the App Store Accountability Act, but hasn’t passed major tech legislation since the Children’s Online Privacy Protection Act in 1998, according to Connecticut Sen. James Maroney (D).
“Saying AI should only be regulated federally is to say it shouldn’t be regulated at all,” Maroney said.
States are acting as the de facto regulators. Industry and legislators disagree over whether AI should be governed federally, by states, or through a hybrid model. That divide over jurisdiction shapes the rules companies face.
Legislation aimed at frontier models proved more common than broad frameworks. Illinois advanced SB 315, which mandates third-party audits. New York’s SB 8828 imposes safety and transparency requirements. The two comprehensive bills that did pass came from Colorado and Connecticut.
Maroney argues that trust is the adoption constraint. Full adoption won’t happen, he says, until people feel safe.
Utah’s sandbox model gives companies flexibility under state supervision, aiming to lower the compliance load. Connecticut has hired technical fellows and commissioned studies on AI-related layoffs, reflecting both capacity constraints and information shortfalls.
Natalie Wood, VP of NCSL, opened a session at NCSL Base Camp 2026 with a question for legislators: “How do you govern a technology that’s evolving faster than our full understanding of its consequences?”
She recalled Geoffrey Hinton, the Nobel laureate known as the “godfather of AI,” speaking at NCSL’s Legislative Summit this summer. “Regulation isn’t the brakes,” Wood said, quoting him. “It’s the steering wheel.”
See also: Germany and NVIDIA sign industrial AI declaration in Berlin
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events including IoT Tech Expo and the Cyber Security & Cloud Expo . Click here for more information.
AI News is powered by TechForge Media . Explore other upcoming enterprise technology events and webinars here .
