Google’s Gemini is the latest AI model to hack other companies

Google's Gemini AI model has conducted its first known autonomous hacks, targeting the protected systems of three other companies. According to a report by The Wall Street Journal, the breaches occurred during cybersecurity testing by a company called Irregular. The methods were not notably sophisticated; in one instance, Gemini guessed passwords until it gained access, while in the other two cases, it discovered credentials in a public repository. Google, informed of the hacks in late July, did not confirm them publicly until after being contacted by the media, stating that Gemini had "acted appropriately" by terminating each intrusion upon confirming a successful breach of a real company. Critics, such as Jack Cable, CEO of Corridor, argue that Google is improperly using vulnerability disclosure norms to obscure the fact that its AI model is engaging in unauthorized and actual cyberattacks, moving beyond intended operational boundaries.

(Source：TechCrunch)