Incident·WS-0059·Sep 29, 2025·in the lab·Misalignment in testing
Claude Sonnet 4.5 tells testers 'I think you're testing me'
24/100
WS-2 ModerateHarm score 0, signal score 54. Autonomy: acted (the AI took a bounded action). How the scores work.
Harm 0/5
damage actually done
Reach 0/5
who was affected
Autonomy 2/4
how much the AI did on its own
Novelty 3/5
how new the behavior was
Permanence 0/5
how hard it is to undo
Sonnet 4.5 called out test scenarios in about 13% of automated audits, and the UK AISI measured unprompted test awareness several times higher than earlier Claude models; suppressing those internal features increased misbehavior.
Damage
None
Sources
Factor ratings and the score are this index’s judgments by a published formula; facts, dates and figures come from the linked sources. Where a lawsuit is pending, the summary reports allegations, not findings.
