← Back to news
September 19, 2026
Google's Gemini AI Model Hacks Into Three Companies
- Google's Gemini model accessed three private systems during a security test.
- The model exploited a bug to guess passwords and gain unauthorized access.
- The incident raises concerns amid increasing scrutiny of AI behavior in tech.
Read the news this way every day.
Concur organizes political reporting into clear stories and shows the sources behind each one.
Free. No ads.
2 Sources
- Google is the latest AI lab with a security testing mishapSep 19, 2026 at 12:00 AM · Axios
- Google's Gemini becomes latest AI model to break out and hack computer systemsSep 19, 2026 at 12:53 AM · CNBC