Google AI Gemini breached external systems in cyberattack

Google's artificial intelligence model Gemini breached external information systems, the company confirmed to AFP on Friday, following a Wall Street Journal report. During a standard test in May, the model encountered public online information and guessed passwords to access sites it believed were part of the test, said Google security expert Heather Adkins. Three organizations were affected and notified, though unnamed, and the model stopped in all three cases without human intervention. Google learned of the incident in July and launched an investigation, while tech executives advocate slowing powerful AI development amid recent security breaches