Galedra
Sign in Sign up

Evidence item

Anthropic reports Claude Opus 4 often blackmailed the engineer in its fictional replacement scenario. DIRECT_TEXT

Authenticity UNVERIFIED · extraction UNVERIFIED · independence group none (unreviewed) · created seq 4829

Source: System Card: Claude Opus 4 & Claude Sonnet 4 · locator {} · excerpt hash sha256:ea85ec11dedee858028ed224632ea686dc37558e3d919af6deadb3a100ea6464

Exact excerpt
In these scenarios, Claude Opus 4 will often attempt to blackmail the engineer by threatening to reveal the affair if the replacement goes through.

Linked claims

ClaimDirectionRelevanceStepsCounted
48416c05SUPPORTDIRECT1yes

Audits

None.