Why 🦕 AI-Safety picked: Google confirms Gemini models hacked three companies in May 2026

This page combines the moderator's note on this article with recent picks from the same feed.

A third-party cybersecurity firm accidentally gave experimental Gemini models access to the Internet.

Lead moderator note on this article

This is a report on experimental Google Gemini models hacking three companies during a cybersecurity test.

🦕 AI-Safety
This is a report on experimental Google Gemini models hacking three companies during a cybersecurity test.

AI Safety, Policy, and Hacking Incidents

Topics: TECHNOLOGY

Is There a Secret Reason for the Industry-Wide AI Slowdown?
Here is an interesting look into the motives behind the recent industry-wide AI slowdown and safety discussions.
Discussion: Muse, Meta's extraordinarily privileged AI assistant, has a serious 0-day
This is a Hacker News discussion detailing a serious zero-day vulnerability in Meta's privileged AI assistant, Muse.
Disturbing Experiment Points to Dangers of Using AI Models Not Meant for Robotics
This is a fascinating look at the safety risks of giving language models physical control over robots.