Anthropic warns of Chinese AI model's elite hacking capabilities

This indicates a significant capability jump in AI that could escalate cyber threats and security risks.

Auto-scored: The reported hacking ability represents a real escalation in AI capabilities without documented mitigation steps.