Cointime update: OpenAI on Nov. 17 said it has partnered with Apollo Research, Goodfire, METR and Transluce to launch a first phase of embedded safety evaluations, giving outside experts access to internal models, source code, training data and monitoring systems. The four firms will assess model alignment and safeguards, with evaluators entering OpenAI offices and using dedicated devices to reach sensitive material. An independent oversight committee under the OpenAI Foundation board will supervise and receive full reports, and OpenAI plans to widen the scope in 2027.
OpenAI Taps Four Independent Firms for Embedded AI Safety Audits
Share on WeChat
Open WeChat → Scan → then tap "…" to send to a chat or Moments.
Tap "…" in the top-right corner to send to a chat or share to Moments.
Related News
41m ago
Alibaba, Tencent and Moonshot AI Limit Offensive Cyber Capabilities in New Models
2h ago
Institutions Cut Data Center Stocks in Q3, Keep Buying AI and Semiconductors
6h ago
Altman Voices Sympathy for Huang Renxun, Says Nvidia Hard to Imagine Without Him
7h ago
Fed Chair Warsh: Staff Modeling Three AI Scenarios for US Economy
8h ago
OpenAI reportedly explored over $5B acquisition of AI ‘mind-decoding’ startup
9h ago
Hard-Tech Firms Including AI Companies Flock to Hong Kong IPOs as Commercial Viability Tests Valuations