OpenAI defends evaluator access terms; ex-researcher urges honoring commitments ASAP

j_asminewang · x · 2026-11-17

Responding to criticism of OpenAI's METR/Apollo engagements, safety lead lamaahmad detailed: evaluators get deep access to some of the most sensitive frontier tools and models; the process will adapt as they learn, with the north star being precision about what's done and how — not a checkbox public report. Authorizing groups of employees to work with evaluators isn't restricting conversations, she says.

Ex-researcher w01fe, while crediting the hard work behind the outcome, pushed back: "we owe it to the world to live up to our commitments ASAP, not at some nebulous future time that might be after the singularity."

Original post →

More from Companies & People

Companies & People channel →