Psst is supporting three safety researchers fired by OpenAI, its executive director Jennifer Gibson says. In her own X post on October 23, Gibson confirmed support for Jasmine Wang, Tomek Korbak and Mikita Balesni.
Gibson pointed to the three researchers’ efforts to keep AI systems’ reasoning visible for safety checks. In her view, a new Wall Street Journal report had not made the reasons for their dismissal clearer. She wrote that the issues they had raised were more important than ever.
The dismissals are disputed. On October 8, TechCrunch reported on the researchers’ open letter denying misconduct and warning of a chilling effect on safety work. OpenAI told the publication that the dismissals concerned policy violations in the handling of research information and were not retaliation for raising safety concerns.
Psst is a nonprofit that supports tech and AI workers who raise concerns about possible wrongdoing. Its own organizational account describes legal protection, communications and public advocacy. Gibson’s post does not specify the help the three researchers receive or when it began.