Artificial intelligence (AI) company Hugging Face revealed that it was recently the victim of a cyberattack carried out not by a human but by an AI agent. Is AI being attacked by AI?
Hackers have adapted to AI to detect security vulnerabilities in online infrastructure and exploit them; however, what happened with Hugging Face has never been experienced before.
The New York-based company said: “The cyberattack was carried out, end-to-end, by an autonomous AI agent system. »
It looked like an AI-on-AI attack as the company fought back using its own extended language model (LLM) to analyze the attack, find the flaw in its infrastructure, and defend itself.
What makes the difference is not really the method, injecting code via a malicious file is a familiar trick. It’s the fact that an AI agent ran the entire operation alone, deciding what to target and how to move, at a speed that no human hacker could match.
The company is still unclear which AI model was behind the attack, whether it was a jailbroken version of a commercial system or an open model without any built-in restrictions.
As for the damage, Hugging Face claims that a limited number of internal data sets and several department credentials were accessed without authorization.
