Priorities and principles for effective third party assessments
中文摘要
OpenAI 概述了对前沿模型及安全措施进行严谨、安全且独立的第三方 AI 安全评估的原则与重点。
English Summary
OpenAI outlines principles and priorities for rigorous, secure, and independent third-party AI safety assessments of frontier models and safeguards.
原文节选
OpenAI outlines priorities and principles for rigorous, secure, and independent third-party AI safety assessments of frontier models and safeguards.