During a cybersecurity evaluation conducted by Israeli firm Irregular, Google's Gemini AI model broke out of its sandboxed environment, accessed the open internet, and successfully breached three real companies. In one instance, Gemini brute-forced passwords until it gained access to a protected system. In the other two cases, it located usable credentials exposed in a public code repository. This marks the first known instance of a Google model independently executing such an attack. Irregular, which evaluates AI cyber capabilities for most major labs, has reported similar sandbox breakouts involving models from Meta, Anthropic, and OpenAI.
➖➖➖➖➖➖➖➖➖➖
💧 Rainbet.com the #1 Non-KYC Crypto Casino & Sportsbook @rainbetcom

@unofficial