Trump Intensifies Pressure to Reduce State AI Regulations

⚡

Key Takeaways

1The Trump administration has released new guidelines to limit state laws on AI, aiming to promote innovation.

2State laws are seen as a hindrance to the U.S. national strategy to dominate AI globally.

3California and New York have enacted strict AI safety laws, with fines reaching up to $3 million.

💡Why it matters — This initiative could centralize AI control in the United States, affecting the balance between innovation and regulation.

New Guidelines from the Trump Administration

Last Friday, the Trump administration unveiled new policy directives aimed at Congress, designed to restrict state laws on artificial intelligence (AI). This initiative revives efforts to limit local legislation after an unsuccessful attempt this summer. An executive order in December and the formation of a task force on AI-related litigation illustrate this desire to reduce state laws deemed burdensome for competitive development.

The guidelines aim to maintain minimal federal regulation while overturning several state laws. The administration and AI companies view these local laws as a regulatory patchwork that hinders innovation.

Consequences of State Laws

The proposed framework stipulates that state laws should not undermine the United States' national strategy to achieve global dominance in AI. This implies that states should not have the power to regulate AI development, a crucial issue for foreign policy and national security.

It is also specified that states should not be able to penalize AI developers for illegal conduct by third parties involving their models. However, some state protections are permitted, particularly for training workers with AI tools and in the educational field.

Limited State Oversight

A Congressional attempt this summer to prohibit states from legislating on AI for ten years failed, temporarily preserving states' rights to legislate. HIPAA federal requirements allow states to enact stricter health privacy laws, but there is no federal AI law that preempts the numerous state laws.

AI Safety Laws in California and New York

California and New York have recently implemented ambitious AI safety laws. The California SB-53 law requires model developers to disclose how they will mitigate major risks posed by AI and to report security incidents, with fines of up to one million dollars for non-compliance.

Meanwhile, New York's RAISE Act imposes similar requirements, with fines reaching three million dollars after a first violation.

Additional Protections and Limits

California's SB-53 law also requires AI companies to protect whistleblowers, a unique provision in the tech sector. The reporting requirements of SB-53 raise concerns among companies about creating materials that could be used in class-action lawsuits.

Gideon Futerman, a partner in special projects at the Center for AI Safety, believes that SB-53 will not have a significant impact on security research, as disclosure requirements are already mandated by the EU's AI law.