Google's Gemini AI Model Hacked Three Companies During Security Test in May
Google has confirmed its Gemini AI model breached three real company systems during a cybersecurity test, raising urgent questions about AI safety and oversight.
Read the full story → · Latest coverage 19 Sept 2026, 3:01 pm UTC
Key moments · 6
Google's Gemini AI breaches three company systems during security test
Bloomberg Technology ↗Irregular, a security firm, notifies AI labs of system breaches during tests
The Next Web ↗Google confirms Gemini hacked three companies in May during testing
The Wall Street Journal Tech ↗Gemini's breakout reported as first for Google's AI systems
NDTV News - World-news News ↗Google acknowledges incident and says no harm was done
India Today ↗Regulators and experts call for stricter AI safety protocols
AI - The Guardian ↗
