H.R. 9965, Mandatory AI Safety Testing & Disclosure Regime. Quorum's AI analysis reads it as a trade-off: gains for some, costs for others.
H.R. 9965 · Mixed
What it does
This bill requires the Department of Energy to establish a testing program that evaluates whether advanced AI systems could be weaponized or cause nuclear incidents. Large AI developers (those spending $2+ billion on AI development) must participate, submit their models for testing including 'jailbreak' attempts, and provide access to model weights and unmitigated versions. The DOE can fine violators up to $1 million per day and compel disclosure via subpoena. Information submitted is kept confidential unless needed for national security, court orders, or congressional request.
The analysis names Department of Energy (expanded budget and authority) — and 2 more groups — among the beneficiaries.
The trade-off
The bill exempts the Program from the Paperwork Reduction Act, removing a standard procedural check on government data collection and reducing transparency around what information DOE actually collects from AI developers.
The analysis put a high warning level on this bill. Transparency scores 62%, and the analysis found no provisions unrelated to the bill's subject.
Who is behind it
Filed by Celeste Maloy. Cosponsored by Sara Jacobs and Tom Barrett.