Collaboration on Adversarial Threats and Security Risks Act

The Collaboration on Adversarial Threats and Security Risks Act aims to clarify the applicability of antitrust laws to collaborations addressing artificial intelligence (AI) security risks. It establishes an exemption allowing non-Federal entities to share information or assistance in good faith for the exclusive purpose of a covered artificial intelligence security purpose. This includes protecting against, identifying, evaluating, or mitigating potential AI threats.

Furthermore, the bill permits non-Federal entities to coordinate or enter agreements to delay or limit the release, deployment, use, or development of AI, specifically to reduce covered artificial intelligence security risks. Such coordination requires prior written notice to the Assistant Attorney General detailing the specific risk and scope of the proposed restriction. These covered risks encompass scenarios where AI could be weaponized by foreign adversaries, facilitate weapons of mass destruction, disrupt critical infrastructure, or become uncontrollable.

The exemption is limited; it does not apply if information is used for non-security purposes or if reasonable internal controls are not in place. While providing an affirmative defense in antitrust actions, the bill explicitly states that it does not permit traditional anti-competitive behaviors such as price-fixing, market allocation, or boycotting. Information submitted for coordination purposes is exempt from public disclosure.

The Attorney General retains the authority to seek injunctive relief against actions that violate antitrust laws, and the exemption does not provide immunity if entities fail to prove their actions were for the exclusive security purpose or if the actions are shown to increase overall AI security risks.