
Anthropic CEO calls for AI model development slowdown to prioritize safety
The story in brief
Anthropic CEO Dario Amodei urged AI companies to slow model development for safety, citing risks like AI being used in multi-agent workflows for reconnaissance and data collection. The firm banned 16 Claude accounts linked to Iranian paramilitary groups, highlighting AI's shift beyond traditional chatbot use. Russia has increasingly relied on intermediaries since Western sanctions followed its Ukraine invasion.