据WSJ报道，黑客利用Anthropic的Claude未经授权进入了OpenAI的系统。

Exclusive | Hackers Used Anthropic’s Claude to Break Into OpenAI WSJ