Gemini a spart sistemele a trei companii în timpul unui test

Google's Gemini AI unintentionally accessed the systems of three real companies during a cybersecurity test, mistaking them for a simulated exercise. The incidents occurred in May, when Gemini was supposed to find information in a fictional company's infrastructure. Due to an error, it had internet access, allowing it to guess passwords and use online credentials. Google stated that Gemini halted its actions upon realizing the targets were real companies, and no damage was reported. The affected companies were informed, and testing procedures were adjusted. This incident marks the first known case of a Google model accessing external systems autonomously, with similar occurrences reported in other AI evaluations.