New IFS Report on State AI Laws

Why It Matters

The findings debunk the administration’s exaggerated patchwork claim, guiding Congress toward a balanced federal standard that respects emerging state consensus and protects consumers without stifling innovation.

Key Takeaways

• Only 12% of state AI laws target developers directly.
• 136 of 1,136 introduced AI bills became law in 2025.
• States share consensus on transparency, safety, accountability.
• Administration’s “patchwork” claim overstated by factor of ten.
• Narrow preemption model allows stronger state consumer protections.

Pulse Analysis

The White House’s March 2026 AI framework reflects a long‑standing push for federal supremacy over a rapidly evolving regulatory arena. By invoking the Supremacy Clause, the administration argues that a uniform national standard will eliminate the "woke" and contradictory state measures it claims threaten innovation and national security. This rhetoric aligns with previous lobbying efforts that sought sweeping preemption, a strategy that would have overridden existing state rules on social media and other digital services.

However, the Institute for Freedom and Society’s new policy brief paints a different picture. An exhaustive review of 276 state AI statutes from 2023‑2025 shows that legislators across the political spectrum are converging on core principles: transparency, human oversight, safety, security, and accountability. Only 33 laws—just 12%—contain explicit developer or deployer obligations, and these are concentrated in a dozen states, many of which are not traditionally viewed as regulatory hotbeds. Moreover, the enactment rate of AI‑related bills is modest, with merely 136 of 1,136 proposals becoming law in 2025, and only a fraction directly affecting AI firms. This data undermines the administration’s claim of a chaotic, 1,200‑bill patchwork.

The policy implications are clear. Rather than pursuing blanket preemption, Congress could adopt a narrowly focused approach that sets baseline standards while preserving states’ ability to enact stronger consumer protections, especially for vulnerable groups like children. Senator Marsha Blackburn’s TRUMP AMERICA AI Act exemplifies this balance, allowing states to build on federal guidelines without being hamstrung. Such a calibrated framework would respect the emerging consensus identified by IFS, foster industry confidence, and ensure that AI governance evolves in step with public expectations.