US SB2937

Establishes federal product liability standards for artificial intelligence systems, holding developers liable under negligence, failure-to-warn, breach of express warranty, or strict liability theories when their AI products cause harm including physical injury, financial loss, emotional distress, or property damage

Creates deployer liability when a deployer substantially modifies an AI system or intentionally misuses it contrary to its intended purpose, with deployers potentially liable as developers if the original developer is outside court jurisdiction, insolvent, or otherwise unable to answer for harm

Prohibits developers and deployers from including contract terms or conditions that waive liability rights, restrict legal forums, or unreasonably limit accountability under the Act, rendering such provisions unenforceable

Grants enforcement authority to the Attorney General, state attorneys general, and private individuals to bring civil actions in federal court seeking injunctive relief, civil penalties, damages, and attorney fees, with a 4-year statute of limitations from discovery of harm

Requires foreign AI developers to designate a U.S.-based agent for service of process before making products available in the United States, with the Attorney General maintaining a public registry and authority to seek injunctive relief against non-compliant foreign developers