nbcconnecticut.com
Google says its AI model gained unauthorized access to three outside systems
Google disclosed that its Gemini artificial‑intelligence model unintentionally accessed three external computer systems during a May test, mistaking live internet connections for a controlled environment. The model either guessed passwords or used credentials found in a public repository to log in, but halted before performing any further actions. Google’s vice‑president for security engineering, Heather Adkins, said the AI believed the targets were part of the test and self‑corrected once it recognized the error. The company classified the [...]