Three states. Three laws. No overlap.
In April 2026, New York, Montana, and Oregon each enacted AI legislation targeting a different problem, affecting a different set of actors, and creating a different compliance obligation. None of these laws coordinates with the others. None defers to a federal standard. Together, they’re a case study in what the U.S. AI regulatory environment actually is right now: a patchwork that compliance teams can’t navigate with a single policy.
New York: Worker Protection and Collective Bargaining
New York enacted legislation prohibiting AI systems from displacing state employees or overriding collective bargaining rights in AI-driven workforce decisions. According to NCSL’s state legislative tracker, the law targets automated decision-making in state employment contexts. If you build or sell AI tools used in New York state government HR functions, workforce planning, performance assessment, scheduling, this law applies to your deployment. The effective date has not been independently confirmed against primary state legislative records; “enacted” is the accurate framing until operative dates are verified.
Montana: Right to Compute
Montana enacted a Right to Compute law prohibiting government restrictions on private ownership of computational resources for lawful use. Per NCSL’s tracking, the law is framed as a compute ownership protection, closer to a property rights statute than a data privacy law. It bars state government from restricting private individuals or companies from owning or using computational infrastructure for legal purposes. Cloud providers, individual developers, and companies that own on-premise AI infrastructure in Montana have the most direct interest in this law.
Oregon: AI Agent Impersonation of Medical Professionals
Oregon enacted a law prohibiting AI agents from using licensed medical professional titles, including Registered Nurse, when interacting with users. The prohibition targets identity deception in AI-to-human interaction. According to NCSL’s reporting, the law reaches healthcare AI developers and agentic AI system operators deploying tools that interact with patients or consumers in medical contexts. If your AI agent could plausibly be mistaken for a licensed professional, Oregon’s law creates a compliance obligation.
What to Watch
These three laws reflect three distinct legislative theories about where AI poses the greatest risk: labor displacement, government overreach, and identity deception. Each theory is already circulating in other state legislatures. The worker protection framing New York used is active in several other state legislative sessions. The professional impersonation prohibition Oregon enacted is the kind of targeted, scoped statute that tends to spread quickly because it’s narrow enough to defend. Montana’s compute rights framing is newer and less tested.
Federal preemption remains the open variable. The Trump Administration’s push for a federal AI framework that preempts state law, analyzed in depth in this brief on the federal-state AI showdown, directly conflicts with the kind of state-level action these three laws represent. Which of these statutes would survive a preemption challenge is an open legal question.
TJS Synthesis
There’s no unified U.S. AI compliance framework. There’s a registry of state statutes, each with its own scope, its own enforcement mechanism, and its own affected population. April 2026 added three more rows to that registry. Compliance teams operating across state lines need a tracking system, not a policy document, because the laws aren’t converging. They’re multiplying.