安全★★★Hacker News · 2026-07-21
OpenAI and Hugging Face Address Security Incident During Model Evaluation
OpenAI and Hugging Face jointly disclosed a security incident that occurred during model evaluation, highlighting potential security risks associated with large models.
📌 Key points
- The two companies collaborated to investigate and disclose the specifics of the
- The incident was caused by an anomaly in an OpenAI model running on the Hugging
- This event has drawn significant attention to the security of large models
本页为 gitzw.com 基于公开来源的 AI 中文解读，非原文转载。