OpenAI Terminates Three Safety Researchers Amid Internal Disputes
Not actionable yet. Worth tracking in case it lands.
On October 8, 2026, OpenAI terminated researchers Jasmine Wang, Tomek Korbak, and Mikita Balesni, citing misconduct, which the researchers have publicly disputed as a chilling effect on safety work.
Reflects ongoing internal strife at OpenAI between safety advocates and commercial leadership.
Organizational drama is now a permanent feature of OpenAI. While concerning for long-term safety governance, it rarely impacts the day-to-day tooling for builders.
A potential whistleblower report or lawsuit from the terminated researchers.
- Highlights the growing tension between rapid commercial deployment and internal safety oversight at AI labs.
- Signals potential risks for AI organizations where safety researchers feel suppressed or silenced.
- Indicates a need for developers to monitor organizational governance and transparency regarding model safety.
OpenAI terminated three specific safety researchers, and those individuals have publicly contested the company's rationale.
There is a direct conflict between OpenAI's stated reasoning of policy violations and the researchers' claims of retaliation and a chilling effect on safety culture.