OpenAI breached by researchers using Anthropic models

Security company hacked into AI lab using tools from its rival