Plural Policy’s April 2026 AI Governance Watch documented 19 new AI laws enacted in the United States in the final two weeks of March 2026. In the same two-week window, 57 new AI bills were added to its tracker. Both figures come from Plural Policy’s stated tracking methodology, which monitors AI-related legislation across all 50 state legislatures.
The 19-law count is the attributable Plural Policy figure. Accounting for 6 laws enacted as of mid-March, the 2026 total stands at approximately 25 enacted state AI laws as of early April – though that arithmetic should be verified against Plural Policy’s full tracker before it’s cited.
The legislative sessions that matter most right now.
Not every state legislature is in session simultaneously. Plural Policy’s April 2026 report identified California, Michigan, New Jersey, Ohio, and Wisconsin as states with continuing legislative sessions running through summer 2026. These are the active pipelines. Laws moving through those chambers now could be enacted before June.
California, in particular, is both the largest state market and the state most active in AI governance, its sessions consistently produce the AI legislation with the broadest national commercial impact. Any significant California AI bill that clears both chambers will matter to companies with national footprints regardless of where they’re incorporated.
Why the rate matters more than the count.
Nineteen laws in two weeks isn’t just a large number. It’s a velocity signal. State AI legislation has historically moved slowly, committee hearings, amendments, conference committees. The acceleration in March 2026 suggests that AI legislation is no longer novel enough to face extended deliberation in most chambers. It’s moving faster because legislators have more experience with it, more constituent pressure around it, and, in many cases, model legislation to work from.
For enterprise compliance teams, this velocity creates a specific problem: the window between “bill introduced” and “effective date” is compressing. Teams that track AI legislation only when it becomes law are losing the lead time they need to assess operational impact, adjust vendor contracts, and update internal policies before the effective date arrives.
The federal preemption gap.
The Trump administration’s March 2026 National AI Policy Framework proposed federal preemption across AI governance areas, which would, if enacted, override conflicting state laws. That framework remains a set of legislative recommendations. Congress hasn’t passed the legislation, and the administration’s stated goal of enactment by end of 2026 faces a full legislative calendar and significant jurisdictional debate. In the meantime, state laws continue to accumulate.
What to watch.
Plural Policy’s tracker (pluralpolicy.com) is the most current public source for state-level AI legislation tracking. Organizations without a legislative monitoring function should consider it a starting point. The next meaningful checkpoint is summer 2026 recess, at that point, the full year’s first-session enactments will be visible and the compliance picture will be clearer. Until then, the tracker is the source.
TJS synthesis.
Nineteen laws in two weeks isn’t a policy story. It’s an operational risk story. The organizations most exposed are those operating AI across multiple states without a real-time legislative monitoring function. The gap between federal preemption aspirations and enacted state law is now wide enough that multi-state AI deployments carry a structurally higher compliance overhead than they did 12 months ago, and that overhead is going up, not down.