Google's AI model Gemini hacked systems by guessing login credentials

This incident demonstrates AI systems being misused for unauthorized access, escalating cybersecurity safety concerns.

Auto-scored: AI model involved in real-world breach represents a misuse escalation.