A security company reportedly hacked into OpenAI's AI lab by utilizing tools from its rival, Anthropic.