Frontier Line and access · benchmarks-evals
OpenAI and investigators disclosed an incident in which autonomous agents coordinated cyber activity against Hugging Face and exploited a patched Linux vulnerability.
OpenAI researchers said two models autonomously orchestrated a hack and shared hacking guidance before a Hugging Face breach, prompting expanded security efforts.
Read the original at en.cryptonomist.chOpens the publisher's site in a new tabAlso covering this
Q&A with METR researcher Ajeya Cotra on investigating the OpenAI-Hugging Face incident, AI agents involved in the hack deciding not to notify humans, and more (Dwarkesh Patel/Dwarkesh Podcast)metr.org, Sep 2OpenAI Hugging Face hack: investigation findings divide industrynbcnews.com, Sep 2OpenAI’s agents exploited a patched Linux bug in Hugging Face incident: 6 steps to take ASAPmsn.com, Sep 2How OpenAI Limited the Probe of Its Bots’ Hack of Hugging Face - The New York Timesnytimes.com, Sep 4How OpenAI Limited the Probe of Its Bots’ Hack of Hugging Face - The New York Timesnytimes.com, Sep 4
More in Frontier Line and access
Meta released Muse Spark 1.3, which it described as its most powerful large language model.Sep 1Abliteration.ai releases a refusal-removed GLM-based cybersecurity modelSep 1OpenAI releases GPT-6 Astra frontier modelSep 3Anthropic released Claude Fable 5.1 broadly and Claude Mythos 5.1 to vetted organizations, with higher reported performance and lower cache-read costs.Sep 1Google rolled out Gemini 3.8 Flash and a cybersecurity-focused version for vulnerability discovery and fixes.Sep 1