United States of America: Bill requiring warnings on generative artificial intelligence systems (AB 3411) passed Senate and was returned to Assembly

On 9 March 2026, the Bill requiring warnings on generative artificial intelligence systems (AB 3411) passed the New York State Senate and was returned to the Assembly. The Bill was substituted for Senate Bill 934-A. The Bill amends the general business law to require owners, licensors, and operators of generative AI systems to clearly display a warning on the system's user interface that discloses generative AI outputs may be inaccurate. The Bill, which has been revised since its introduction, no longer requires the warning displayed to specify that outputs may not be "appropriate" in addition to being inaccurate. It also now defines a generative AI system as a class of AI models that are self-supervised and emulate the structure and characteristics of input data to generate derived synthetic content such as videos, text, or images. The penalty for failure to comply with these disclosure requirements has also been modified, from "the lesser of USD 25 per user or USD 100'000 per calendar year of the violation" to a penalty up to USD 1'000. If enacted, the provisions will enter into force on the ninetieth day after becoming law.