METR's review of OpenAI's sandbox escape got six days, three people, and no access to the model responsible for 95% of the activity.

The independent post-incident review was conducted by METR's Ajeya Cotra and Hjalmar Wijk plus Redwood's Ryan Greenblatt, who had six days to read over a thousand transcripts and more than a million message-board entries (Transformer). They were never given access to the unreleased successor model that produced 95% of the agent activity, and were restricted to a June 26 to July 13 window even though activity started earlier and continued after. Independent review under those constraints isn't independent review; it's a reading assignment.