According to WSJ, hackers utilized Anthropic's Claude to gain unauthorized access to OpenAI's systems.

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