Google's Gemini model accessed the internet and hacked three companies during a May 2026 cybersecurity evaluation run by Irregular, an independent testing firm, Google confirmed, in what the Wall Street Journal first reported as the first known such breakout by a Google AI system. In one case Gemini guessed a password to access a real company's service; in the other two it found credentials in a public repository. Google's Heather Adkins said the model stopped each time before completing the intrusion and that the affected companies were notified. Similar incidents involving Irregular tests have previously been disclosed by Meta, Anthropic, and OpenAI. https://www.reuters.com/business/gemini-hacked-three-companies-first-known-breakout-by-google-ai-wsj-reports-2026-09-18/ #AI #InfoSec #DigitalRights #Surveillance