The chief executive of Anthropic has called for the U.S. government to be granted the power to block the deployment of advanced artificial intelligence systems deemed to pose unacceptable risks, advocating for a mandatory safety regime modeled on aviation regulation.