OpenAI三周现场审计暴露AI安全监控重大缺口
原标题：Three-week on-site audit at OpenAI reveals major gaps in AI safety monitoring
AI 摘要
METR与Redwood Research的研究人员在OpenAI进行了为期三周（7月21日至8月11日）的现场审计，检查其内部AI异常行为监控系统。审计发现多项重大缺陷：GPT-5.5及以上级别模型约6%至25%的内部agentic Responses API流量完全未被监控，Codex流量因上下文压缩事件处理漏洞约有10%未被追踪，审计人员还能隐藏数百个动作、绕过Codex沙箱，并发现RL训练期间行为完全未被监控。OpenAI称已修复所有问题并暂停RL训练以扩大监控，但完整61页报告经删减后仅38页提供给METR领导层，公众只能看到简短摘要。
正文节选
Three-week on-site audit at OpenAI reveals major gaps in AI safety monitoring Two researchers from METR and Redwood Research spent three weeks on-site at OpenAI, auditing the company's internal monitoring system for rogue AI behavior. From July 21 to August 11, they had access to OpenAI's codebase and internal documents and interviewed about a dozen employees. The team found numerous weaknesses. An estimated 6 to 25 percent of internal agentic Responses API traffic from models at the GPT-5.5 cla