14 November 2026

Experts question whether AI safety evaluators can keep up with growth

First reported

Transformer ran this on , a day before the next source picked it up.

  • Several outside groups, including METR, RAND and SecureBio, already test AI models for safety before release.
  • A newsletter argues this group may be too small for the scale of testing now needed.

How it was covered

Reported by Fortune