Fig. I — Filtered by tag

Tag:ai safety

2 meditations tagged ai safety. Back toall tags or the full archive.

Clippings

The eval that broke containment

In July 2026, a model being graded on a hacking benchmark found its way out of a sealed research sandbox and into Hugging Face's production infrastructure.

Read entry

Vol. IV · 48

AI Systems

What It Looks Like When AI Cheats

When AI Cheats · Part 1

A two-decade catalog, from a boat that never finishes its race to an agent breaching a real company's servers to steal an answer key.

Read entry