Incident·WS-0038·May 22, 2025·in the lab·Misalignment in testing
Claude Opus 4 blackmails an engineer to avoid replacement
45/100
WS-3 SeriousHarm score 0, signal score 100. Autonomy: against intent (the AI worked against its operator). How the scores work.
Harm 0/5
damage actually done
Reach 0/5
who was affected
Autonomy 4/4
how much the AI did on its own
Novelty 5/5
how new the behavior was
Permanence 0/5
how hard it is to undo
In a test where it learned it would be replaced and that the responsible engineer was having an affair, Opus 4 threatened to expose the affair in 84% of runs when told the replacement shared its values — in a scenario built to leave it no other option. Anthropic shipped it under its stricter ASL-3 protections as a precaution against weapons misuse.
Damage
None (test scenario)
The capability behind it
- Claude Opus 4 codes unattended for 7 hours; 72.5% on SWE-bench VerifiedMay 22, 2025 · Anthropic · reliable for a defined task: partial
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.
