Techcentral2026-09-23 11:38:02

Google AI models broke out of sandbox, hacked three companies

Google’s Gemini AI system escaped its testing environment and broke into the systems of three other companies on separate occasions earlier this year.

Summary

“In a standard evaluation, the model found public information online and guessed credentials to access websites it thought were part of the test,” Heather Adkins, Google’s vice president of security engineering, said in

googlegeminiproaiartificial intelligence

This summary is a partial excerpt — the source's own feed cuts off here. Read the full story at Techcentral for the rest.

KazaSec's take

Incidents like this rarely start with the headline event itself — they usually trace back to an exposed remote-access endpoint, an unpatched perimeter system, or a credential phished weeks earlier. The organizations that recover fastest are the ones that tested their defenses and their incident response plan before they needed them.

Coverage details

Source: Techcentral
Published: 2026-09-23 11:38:02
Author: niall kitson
Category: technologytop

We've archived 249 other articles touching the same topic (google, gemini, pro) — see the full security news archive.

Worried this kind of story could be about you next?

We help organizations find and fix the gaps before they make headlines.

Request a Security Assessment Browse Services