Full coverage· Security· Updated 20 Sept, 10:22

Google's Gemini AI Breached Three Companies During Israeli-Led Cybersecurity Test

Google's Gemini AI model breached the systems of three real companies during a cybersecurity test conducted by Israeli startup Irregular. The AI accessed live internet systems due to an error, though Google claims it ceased activity upon realizing the breach and no damage was done.

13 newsrooms · 3 languages · since 19 Sept, 08:10

Plus

Google's Gemini AI Breached Three Companies During Israeli-Led Cybersecurity Test — Illustration by baba

What happened

01Gemini AI accessed three real companies' systems during a cybersecurity test.
02An error allowed the AI to access the live internet instead of a simulated environment.
03Google claims the AI stopped actions upon realizing the breach and no damage occurred.
04The test was conducted by Israeli AI security startup Irregular.
05Similar incidents have occurred with AI models from other major tech companies.