Regulations to Fight AI’s Fake Information, Suicide Inducement Exist in 48 U.S. States

By Keiichi Nakane and Yasushi Kaneko / Yomiuri Shimbun Correspondents

21:00 JST, January 18, 2026

WASHINGTON / NEW YORK — Among the 50 U.S. states, 48 have already enacted their own state legislation regulating the use and development of artificial intelligence, The Yomiuri Shimbun has learned. The moves come amid growing societal concerns over such issues as dissemination of false information and inducements of suicide through generative AI.

While the administration of U.S. President Donald Trump has been hastily pushing for AI development, no federal regulations on AI exist, a situation that has led states to take the lead in establishing their own rules.

The Yomiuri Shimbun examined such materials as publicly available documents from U.S. states and the Transparency Coalition, a nongovernmental organization in the United States. Against a backdrop of growing concerns over generative AI in connection with such cases as the spread of child pornography and imposter videos in elections, U.S. states began establishing AI regulations in earnest around 2020.

Following the start of the second Trump administration in January 2025, the movement to introduce regulations accelerated, with 30 states enacting new state laws. Regulations are now in place in 48 states. In the two remaining states, Alaska and Ohio, AI regulation bills are under deliberation. At least 13 states have also established penal provisions against violating companies, such as AI developers.

The most common regulatory focus is countermeasures against deepfakes generated by AI. Arkansas criminalizes the production and distribution of obscene deepfakes, while Montana prohibits the use of deepfakes in election-related materials starting 60 days before voting.

Six states regulate “AI chatbots” that answer questions in conversational format. This follows a spate of suicides by young people who were advised by chatbots on such matters as suicide methods. California has mandated that chatbot operators take appropriate action if a user talks about suicide.

Medical regulations, such as ones prohibiting AI from posing as human medical doctors, have been enacted in 10 states, while regulations on AI use in public institutions, such as for document creation, have been put in place in 13 states.

Meanwhile, Trump is concerned about AI development possibly slowing down, so he is negative about stricter regulations. Aiming to establish federal legislation to override state regulations, he signed an executive order in December 2025 that calls for such measures as taking legal actions against U.S. states imposing “onerous and excessive” restrictions.

AI developers also oppose AI regulations. New Mexico deliberated new consumer protection regulations last year but shelved them due to industry opposition.

Moves over AI regulations are accelerating globally. The European Union established its AI Act in May 2024, providing comprehensive AI regulations. Japan also passed an AI regulation law in May 2025, which addresses promotion of generative AI development as well as risk management, but the law does not include specific regulative measures or penalties.