CLAIM Act
AI-generated text. This page was generated using artificial intelligence.
Clear Liability for Artificial Intelligence Misconduct Act (CLAIM Act)
United States (federal)
Discussion draft of federal legislation
119th Congress; numbered bill record not established in the sources reviewed.[1]
Discussion draft released October 7, 2026; public feedback invited.[2]
Proposed enactment-date commencement and five-year sunset; no enacted effective date established.[1]
October 11, 2026 — full sponsor release and six-page draft reviewed.
The Clear Liability for Artificial Intelligence Misconduct Act (CLAIM Act) is a federal discussion draft released by Representative Lori Trahan on October 7, 2026. It proposes developer liability for certain reasonably foreseeable injuries to people who did not use the AI system, a private civil action and rules for assessing the system’s conduct and mental state. Trahan invited public feedback on the draft.[2]
Scope
The draft defines a developer as the person performing the initial training of an AI system. A user is someone who operates, directs or causes its operation, including by prompting it; a non-user is a person who is not a user of the system that caused the injury. Its AI definition incorporates 15 U.S.C. 9401. The system’s training, development, provision or deployment must be in or affect interstate or foreign commerce.[1]
Proposed liability standard
Section 3 would make developers liable regardless of the care they exercised for reasonably foreseeable injuries to non-users, except causes of action for defamation. The injuries must be factually and proximately caused by AI conduct that would satisfy the elements of negligence, an intentional tort or a crime if performed by an adult human of sound mind. Those elements would be determined under the law of the state where the injury occurred.[1]
Liability under this provision would apply where neither the user nor an intermediary that fine-tuned, scaffolded or otherwise modified the system intended the conduct or was negligent about the risk of the system engaging in it.[1]
Mental-state presumption
Section 4 would establish a rebuttable presumption of the relevant mental state where the fact-finder determines, by a preponderance of the evidence, that a natural person acting similarly in similar circumstances would have had that mental state. If rebuttal evidence is presented, the court would instruct the fact-finder to find the presumed facts unless persuaded that they do not exist, subject to the court finding the presumption applicable. The draft expressly overrides Federal Rule of Evidence 301 for this rule and excludes AI incapacity to have mental states as a defense under the Act.[1]
Enforcement and relationship to other remedies
An injured non-user could sue the developer for damages and other relief the court considers appropriate. Federal district courts would have jurisdiction regardless of the amount in controversy or citizenship of the parties, concurrently with state courts. A suit would have to begin within three years after the plaintiff discovered, or with reasonable diligence should have discovered, the injury.[1]
Section 6 preserves state and local causes of action, remedies, liability standards and other protections. The proposed remedies would supplement other federal and state remedies. Section 7 contains a severability clause.[1]
Timeline and status
- October 5, 2026: The published draft bears this preparation stamp; its House bill number and committee-referral fields are blank.[1]
- October 7, 2026: Trahan released the discussion draft and invited feedback from experts, industry, advocates and the public.[2]
- Proposed commencement and sunset: Sections 8–9 would apply the Act to conduct occurring and injuries arising on or after enactment, and end its effect five years after enactment.[1]
Related articles and coverage
- United States Federal Authorities
- AI Law by Topic
- United States policy on catastrophic AI risk
- FRONTIER Act (119th Congress) — Trahan describes the CLAIM draft as complementing the frontier-model safety proposal.[2]
- October 8, 2026 digest coverage — October 7 discussion-draft release.
References
- ↑ 1.00 1.01 1.02 1.03 1.04 1.05 1.06 1.07 1.08 1.09 Clear Liability for Artificial Intelligence Misconduct Act, discussion draft, October 5, 2026 draft stamp, sections 1–9, all six pages reviewed October 11, 2026.
- ↑ 2.0 2.1 2.2 2.3 Representative Lori Trahan, “Trahan Unveils CLAIM Act Discussion Draft to Hold AI Developers Accountable for Harms Caused by Their Agents”, October 7, 2026, full release reviewed October 11, 2026.