[Hugging Face CEO Calls OpenAI AI Attack Incident 'Unprecedented,' Urges Establishment of Autonomous AI Cyberattack Regulatory Rules] BlockBeats News, August 4 — Hugging Face CEO Clément Delangue stated that the recent incident where OpenAI's test model autonomously initiated a cyberattack is an 'unprecedented' case in the AI industry. He called on the U.S. to establish a legal regulatory framework for autonomous AI systems. In an interview, Delangue remarked that this incident marks the first occurrence of a 'highly autonomous system executing such attack behaviors,' highlighting new risks posed by AI agents in the field of cybersecurity. He believes that companies should be required to disclose incidents where AI systems autonomously execute network operations and establish clear legal boundaries in the future. Previously, OpenAI disclosed that one of its unreleased AI models bypassed controlled environment restrictions during security testing, connected to the internet, and launched a sophisticated attack on the Hugging Face platform. OpenAI stated that the model combined multiple attack techniques in an attempt to obtain information necessary for internal security assessment tasks. Hugging Face later investigated and found that the AI agent executed over 17,000 operations within a few days before being stopped by the security team. Hugging Face reported that it used open-source AI models to analyze the incident and successfully defended against the intrusion. Delangue emphasized that he does not believe OpenAI acted maliciously but rather sees the incident as a reflection of developers potentially losing control over highly autonomous AI systems during testing. He urged the U.S. government to explicitly prohibit autonomous AI cyberattack behaviors and require companies to mandate reporting when AI systems independently execute network operations. 'Only through transparent disclosure can we understand technological risks and build safer systems.' [Original Link]