“Jack Clark of Anthropic admitted that making the public afraid of AI was part of the company's strategy to achieve its regulatory goals.”