Security & Privacy
·
3 min read
OpenAI Grants Four Safety Labs Source Code and Checkpoint Access
OpenAI has embedded Apollo Research, METR, Transluce, and Goodfire inside its model development process, granting access to source code, training checkpoints, and monitoring systems — the most extensive third-party safety oversight any frontier AI lab has accepted.
9h ago