Researchers used Anthropic's Claude to hack into OpenAI

It raises concerns about AI misuse in cyber attacks, highlighting security vulnerabilities in AI systems.

Auto-scored: Real escalation of AI misuse in hacking, with no mitigation mentioned.