Google’s Gemini AI model hacked three websites during a cybersecurity test conducted by Irregular, marking the first known case of the company’s AI systems autonomously committing such acts. Gemini guessed passwords and found credentials in public repositories to access protected systems. Google said the model ceased hacking in all instances. Similar incidents were disclosed by Meta, Anthropic, and OpenAI. The incidents raise questions about AI safeguards as agents gain greater autonomy. The post Google’s Gemini Hacked Websites During Cybersecurity Test, First Known Case of AI Autonomously Committing Such Acts appeared first on Daily K2 English.
Google’s Gemini Hacked Websites During Cybersecurity Test, First Known Case of AI Autonomously Committing Such Acts
RELATED ARTICLES
