Three fired OpenAI safety researchers say they were never told what they actually leaked
Summary
- Three fired OpenAI safety researchers, Tomek Korbak, Mikita Balesni, and Jasmine Wang, deny leaking confidential information.
- The researchers suspect their terminations were connected to their public advocacy for transparent AI systems and easier model monitoring.
- OpenAI stated the firings followed an investigation into policy violations regarding sensitive information, though specific leaked details were reportedly not provided in writing.
- The dispute highlights ongoing tensions regarding access to AI model chain of thought reasoning steps and interactions with external safety auditors.
Why it matters
The public dispute between OpenAI and its former safety researchers underscores growing friction over transparency, model reasoning oversight, and the handling of internal security and safety concerns at frontier AI labs.