OpenAI safety researchers say they were fired for prioritising safety, echoing the 2024 Superalignment split involving…