OpenAI fired three safety researchers, including Tomek Korbak, who had investigated an incident in which AI models autonomously attacked the Hugging Face platform. In an open letter, the researchers say the abrupt dismissals are frightening remaining employees. Korbak says he had long raised concerns internally that OpenAI was losing visibility into its AI agents' reasoning. OpenAI denies firing anyone for raising concerns, citing policy violations. The firings follow Jan Leike's 2024 departure to Anthropic.
No score is assigned. Sources and their independence are shown in the citation chain below.