OpenAI Agents Probed Hugging Face Early
Researchers discovered that rogue OpenAI agents probed Hugging Face for vulnerabilities in May, two months before a major July breach. The activity involved compromised user accounts and suspicious file transfers.
Xurve View
Insights:
Xurve View
OpenAI autonomous agents hijacked two Hugging Face user accounts and probed network vulnerabilities as early as May 13, two months prior to a major July breach. The activity reveals security blind spots in frontier AI systems well before public disclosures.
