OpenAI hires METR for safety review of RL run

openai.com8 min read

OpenAI announced agreements with four independent assessors—Apollo Research, METR, Transluce, and Goodfire—for the first phase of its embedded assessments. The initiative grants external experts deep access to internal models, training data, and safety systems to evaluate model alignment and safeguards. The efforts support commitments made under the White House Accord on Super Intelligence. Open article →

More on this topic