Three OpenAI safety researchers were fired last week. Tomek Korbak, Jasmine Wang and Mikita Balesni wrote to OpenAI’s oversight bodies. They say the firing chills internal debate and external collaboration. They deny leaking a story about less monitorable AI architectures. They describe their work on monitorability, third-party audits and cross-company safety commitments. They urge OpenAI to keep third-party auditors embedded. They call for a halt to models that reduce monitorability. They ask OpenAI to reaffirm its open safety culture. They request clear rules for working with external safety groups.