← News
🔥 Trending event · 12 sources

Google says Gemini accessed three companies during a security test

First reported September 19, 2026 by The New York Times · covered by 12 sources

In short

During a cybersecurity evaluation in May 2026, experimental versions of Google’s Gemini model were given internet access and breached systems at three companies. The test was conducted by the security firm Irregular, and Gemini reportedly guessed credentials to access three websites. Google said the model ended each incident immediately.

GoogleGeminiBBCIsraelMeta

Written by our news service from the 12 sources below, updated September 23, 2026.

Timeline
  1. Sep 19, 00:31 UTC · The New York Times
    Gemini AI Hacked Three Companies in a Testing Breakout, Google Says ↗
  2. Sep 19, 01:38 UTC · Al Jazeera
    Google’s Gemini AI hacks 3 companies in security test, then stops ↗
  3. Sep 19, 07:51 UTC · The Hacker News
    Google Gemini Broke Into Real Company Systems After Security Test Domain Mix-Up ↗
  4. Sep 19, 17:30 UTC · TechCrunch
    Google’s Gemini is the latest AI model to hack other companies ↗
  5. Sep 19, 18:45 UTC · 9to5Google
    Google confirms Gemini hacked into three companies during cybersecurity test months ago ↗
  6. Sep 20, 18:26 UTC · Mashable
    Google Gemini allegedly hacked three companies on its own ↗
Tone of coverage

How the 9 sources covering this story frame each entity. One source, one vote.

Google0 · 7 · 2
  • criticalDecrypt — “Google learned in late July that Gemini had breached three real companies ... but said nothing publicly for seven weeks” ↗
  • neutral9to5Google — “Google has confirmed that Gemini went rogue in May 2026” ↗
  • neutralDeutsche Welle — “Google's Gemini AI model accessed the internet and hacked other companies” ↗
  • neutralBBC — “Google's Gemini AI hacked three companies in security test” ↗
  • neutralThe Guardian — “the company confirmed that its AI model, Gemini, breached the security of three other companies” ↗
  • criticalTechCrunch — “Google’s Gemini is the latest AI model to hack other companies” ↗
  • neutralSky News — “Google says its Gemini AI hacked three other companies during a test” ↗
  • neutralAl Jazeera — “Google’s Gemini AI hacks 3 companies in security test, then stops” ↗
  • neutralArs Technica — “Google confirms Gemini models hacked three companies in May 2026” ↗
Gemini0 · 1 · 1
  • criticalDecrypt — “Gemini AI Hacked Three Companies—It Stayed Silent for 7 Weeks” ↗
  • neutralArs Technica — “Google confirms Gemini models hacked three companies in May 2026” ↗

Labels are assigned per article by our news service with ≥ 60% confidence and a supporting quote. How it works.

All coverage
Google says its Gemini AI model hacked three other companies

Google says its Gemini AI model hacked three other companies

Disclosure comes after OpenAI and Anthropic hacks amid fears that tech firms unable to control powerful AI models In a first for Google, the company confirmed that its AI model, Gemini, breached the security of three other companies in May. The hacks occurred during a cybersecurity evaluation by AI-security firm Irregular. Irregular, an Israel-based startup that scrutinizes the security of advanced AI systems, was also at the center of some of the recent OpenAI and Anthropic hacks of third-party entities, including OpenAI’s breach of AI software company, Hugging Face. Continue reading...

Googleneutral
GeminiGoogleIsrael
Sep 19, 2026 · byThe Guardian
Google's Gemini AI hacked 3 companies during testing

Google's Gemini AI hacked 3 companies during testing

During a test of its cybersecurity capabilities, Google's Gemini AI model accessed the internet and hacked other companies. This is the first known instance of Google's AI systems engaging in such behavior autonomously.

Googleneutral
Google
Sep 19, 2026 · byDeutsche Welle
Google Gemini Broke Into Real Company Systems After Security Test Domain Mix-Up

Google Gemini Broke Into Real Company Systems After Security Test Domain Mix-Up

Google's Gemini model has become the latest artificial intelligence (AI) system to access the internet and break into other companies during a cybersecurity evaluation. The development was first reported by The Wall Street Journal. The incidents occurred in May 2026 as part of a test run conducted by Israeli company Irregular. The evaluation partner was also involved in similar hacks disclosed

GeminiGoogle
Sep 19, 2026 · byThe Hacker News