Sen. Gounardes Proposes Major New AI Safety and Transparency Regulations

The release states that the bill builds on Gounardes' Responsible AI Safety and Education Act, referred to in the release as the RAISE Act. The press release appears under a 2026 date in the senator's newsroom.

According to the release, the legislation would establish independent third-party analysis and assessments of existing artificial intelligence models. The release also states that the bill includes a "kill switch" that could be applied if a model is deemed to pose a catastrophic risk. The word "independent" is used in the release to describe that analysis and assessment. The release does not name the entity that would carry out the third-party assessments, does not specify which models or developers would fall within scope, and does not state the threshold at which a model would be judged to present a catastrophic risk.

The text as published does not give a bill number, a committee referral, a timeline for consideration or an effective date, and no further provisions of the bill are described. The release describes the measure as strengthening safety and transparency standards for advanced AI without defining which systems count as advanced AI.