California advances landmark legislation to regulate large AI models

Associated Press

SACRAMENTO, Calif. (AP) — A California landmark legislation to establish first-in-the-nation safety measures for the largest AI systems has cleared an important vote. The bill advanced Wednesday would require AI companies to test their models and publicly disclose their safety protocols so they can’t be potentially manipulated to wipe out the state’s electric grid or help build chemical weapons. Supporters say it would set much-needed safety ground rules for large-scale AI models in the United States. Opponents said the proposal would hurt the industry. The proposal would only apply to the largest systems that are still being developed. It still needs a final Senate vote before reaching the governor’s desk.