Unanswered question on the OpenAI "Nightingale" incident: deployed models or internal research?

S_OhEigeartaigh · x · 2026-09-04

Researcher S. Ó hÉigeartaigh asks whether the reported OpenAI/Nightingale incident involved deployed models or internal research models — a detail he can't find anywhere. The distinction matters greatly for assessing severity, and OpenAI's disclosure remains unclear.

Original post →

More from Safety

Safety channel →