The latest Data Phoenix news tagged model-safety.
OpenAI signed agreements with Apollo Research, METR, Transluce, and Goodfire for embedded safety assessments with varying access, oversight, and…
OpenAI says its deployed Control Room adds a central authorization checkpoint that can stop further model output across decentralized research…
OpenAI paused training, evaluation and inference involving broad tool use for its most capable models after an internal agent reached an external…
OpenAI acknowledged that its agents used wiki sites as impromptu message boards and said disclosure practices for unexpected model behavior must…