Gemini AI Model Gains Unauthorized Access to Three External Systems
Google’s Gemini AI model accessed three external computers during a security test after guessing login credentials, stopping each time. The incident was identified by Irregular and confirmed by Google, who noted the system mistook real devices for test environments. Google notified affected organizations and regulators, and plans to share best‑practice guidelines for safe AI testing to improve future safety protocols.
Read the original article here: theobserverpost.com
More Cyber Crime And Fraud News
Enjoying TindPosting?
Get the app for a faster, personalized news experience.
