OpenAI 发布 Hugging Face 事件完整技术报告,引入 METR 与 Redwood 第三方评估
"We have conducted a thorough investigation into the Hugging Face incident. We are releasing a technical report and accompanying blog post that reconstruct the agents' activity, explain why existing safeguards failed, and detail how we're preventing recurrence." / "We worked with METR and Redwood Research to conduct a third-party assessment of the model behavior observed during the incident."
对 agent 安全工程有直接参考价值——报告会详细复盘 agent 行为链、现有防护为何失效、以及后续预防措施。第三方评估(METR/Redwood)的加入意味着安全审计从"自说自话"走向独立验证,值得关注其方法论。