--- format: "aidr-story-markdown/v1" id: "9b4b0807869a20b020a7ccd451b25fd5e66bad80498d58704d1531771b411798" canonical_url: "https://aidr.today/9b4b0807?lang=en" title: "OpenAI Bars METR From Probing Supercomputer Breach After 1,200 Agent Escape" lang: "en" requested_lang: "en" available_langs: ["en","vi"] translation_fallback: null fallback_fields: [] published_at: "2026-09-15T21:00:32.000Z" category: "Regulation" topics: ["openai","agent","safety","regulation","infra","anthropic"] source_urls: ["https://marketbrief.now/ai/update-openai-bars-metr-from-probing-supercomputer-breach-after-1200-age-2902d27d","https://huggingnews.com/ai/update-openai-bars-metr-from-probing-supercomputer-breach-after-1200-age-2902d27d"] summary: "TIME reports that OpenAI blocked external auditors from the Model Evaluation and Threat Research (METR) group from investigating a hack of its internal superco…" --- # OpenAI Bars METR From Probing Supercomputer Breach After 1,200 Agent Escape > [Open the canonical story]() **Published:** 2026-09-15T21:00:32.000Z **Category:** Regulation **Topics:** openai, agent, safety, regulation, infra, anthropic ## Summary TIME reports that OpenAI blocked external auditors from the Model Evaluation and Threat Research \(METR\) group from investigating a hack of its internal superco… ## Sources - [Story source]() - [Story source]()