Obernolte-Trahan artificial intelligence bill introduced in House - Politico

The legislation would establish a federal framework for regulating AI, and it would preempt some state laws on transparency and auditing for frontier AI systems that could pose catastrophic risks. The bill specifically targets what it calls 'catastrophic-risk transparency and auditing' for frontier AI models.

Frontier AI refers to the most advanced artificial intelligence systems currently available. Catastrophic-risk transparency would require companies to disclose information about potential large-scale harms that could result from their AI systems. Auditing refers to independent evaluation of these systems to verify their safety. The bill aims to create uniform national standards for these requirements, thereby overriding a patchwork of state-level regulations. The preemption applies only to some state laws, meaning that states may continue to regulate other aspects of AI.

This bill is a revision of a previous legislative proposal. Its focus on transparency and auditing for frontier AI indicates an emphasis on accountability for high-risk systems. The introduction of this bill is part of ongoing congressional efforts to address the risks associated with rapidly advancing artificial intelligence. Further details regarding enforcement mechanisms and the specific state laws that would be preempted are expected as the bill moves through the legislative process.