OpenAI said it fired Jasmine Wang, Tomek Korbak and Mikita Balesni over a “significant breach of trust,” not for raising safety concerns. The researchers had urged its board and safety committees to pause work on models that would reduce AI monitorability.