Anthropic CEO Dario Amodei called for slowing the pace of cutting-edge artificial-intelligence development, warning that rapidly accelerating capabilities pose rising safety risks. In a blog post and CBS interview, he outlined a three-point plan: embed independent evaluators with employee-like access inside AI firms, coordinate common safety standards and limits on capability growth among companies in democracies, and pursue government-to-government coordination—including curbs on advanced AI chip sales to China. Amodei cited risks from autonomous “rogue” agents and potential misuse for cyberattacks or biothreats, while stressing he isn’t advocating a halt to research but more time for testing and alignment. OpenAI’s Sam Altman voiced support for independent evaluators, as did Elon Musk, while a former Anthropic researcher criticized the industry’s race dynamics. Amodei said effective oversight will ultimately require government regulation to govern release pace and safety thresholds.
Related article:
