The CEO of Anthropic is calling for government authority to block high-risk ai models. This suggestion implies a desire for regulatory oversight of ai development. The specifics of what constitutes a high-risk model and how such authority would be implemented are not clear.