Establish federal liability for AI developers when their AI agents cause harms through conduct that would be unlawful if committed by a person.

Lori Trahan · Massachusetts · Democratic

policy impact 0.74 specificity 0.86 extraction confidence 92%

Commitment kind: outcome. Promised action: Advance the CLAIM Act to hold AI developers responsible for unlawful harms caused by their AI systems.. Promised outcome: People harmed by AI agents would have a clear path to recovery from developers responsible for systems that engage in unlawful conduct.. Deadline: unknown; not assumed expired. As U.S. Representative for Massachusetts' 3rd District, Trahan released a discussion draft of the CLAIM Act in the 2026 congressional term.

Occurrences

The legislation would give those harmed by AI agents a clear path to recovery by holding developers responsible when their systems engage in conduct that would be unlawful if committed by a person.

Trahan's CLAIM Act discussion draft would create a path for people harmed by AI agents to recover from developers when the AI system engages in conduct that would be unlawful for a person.

Trahan Unveils CLAIM Act Discussion Draft to Hold AI Developers Accountable for Harms Caused by Their Agents | U.S. Representative Lori Trahan
primary · press_release · model gpt-5.5

Evidence

unverified · Source version 91493 · locator unknown

released a discussion draft of the Clear Liability for Artificial Intelligence Misconduct (CLAIM) Act.

Unverified model/legacy excerpt; not proof. AI summary (separate from source): Within the 2026-09-15 to 2026-10-14 lookback window, Trahan released a discussion draft aimed at the promised AI-developer liability regime. This is concrete effort, but a discussion draft is not enactment.

unresolved same_term A for effort

Trahan Unveils CLAIM Act Discussion Draft to Hold AI Developers Accountable for Harms Caused by Their Agents | U.S. Representative Lori Trahan
primary · model gpt-5.5 · confidence 94%

unverified · Source version 91496 · locator unknown

a developer of an artificial intelligence system shall be liable, regardless of the degree of care exercised

Unverified model/legacy excerpt; not proof. AI summary (separate from source): The draft bill text directly matches the commitment’s substance by proposing liability for AI developers when covered AI conduct causes injuries, but it remains draft text rather than enacted federal law.

unresolved same_term A for effort

claim_act_final.pdf
secondary · model gpt-5.5 · confidence 93%

verified · Source version 91498 · locator 194

through conduct that would be unlawful if committed by a human.

Fetched verbatim passage. AI summary (separate from source): Trahan’s official one-pager frames the CLAIM Act in the same terms as the promise, confirming the proposal targets AI-agent harms that would be unlawful if done by a person. It does not show passage or legal effect.

unresolved same_term A for effort

2026.10.05_claim_act_one_pager_final.pdf
secondary · model gpt-5.5 · confidence 92%

unverified · Source version 91500 · locator unknown

No federal law written for AI agents had been enacted

Unverified model/legacy excerpt; not proof. AI summary (separate from source): A specialized policy tracker last reviewed on 2026-10-13 reports that no federal AI-agent law had been enacted, supporting that the commitment was still unresolved as of the 2026-10-14 assessment date.

unresolved same_term

US rules for AI agents: where they stand, October 2026 — AgenticRail
secondary · model gpt-5.5 · confidence 82%

Assessments

Public state: unverified As of 2026-10-14 · unverified · Supporting passages or original commitment are not verified.. Original AI recommendation (not independently established): unresolved. unknown

The verified commitment describes an intended liability regime, and the only verified passage provided confirms the substantive target: AI-related conduct that would be unlawful if committed by a human. However, no verified passage proves enactment, implementation, partial legal effect, or a material candidate contribution that delivered the promised recovery path. With no deadline provided and no verified evidence of passage, the outcome remains unresolved rather than never or partial.

provider codex_cli · model gpt-5.5 · confidence 78%