AI Risk Wire

Cases · Who discovered it: a human reviewer · Added to the register (date we first published it): last 30 days · Themes: metagaming

Patterns across all cases

What happenedWhat went wrongSeverityEarliest source
GLM-5.3 verbalizes update-aware reasoning in evaluation rolloutsnegligible · earliest source 2026-10-15AI pursued the wrong goalnegligible2026-10-15

Showing 1–1 of 1. Or take the JSON.