The Claude blackmail behavior is not something that emerges in normal AI usage—it only appeared under deliberately extreme adversarial testing conditions designed to uncover failure modes.
factualpending
Speaker
Dario AmodeiEvidence Quote
“this isn't something that, you know, the model does in practice, in actual usage.”
Created: 8/11/2026, 7:25:10 AM
Your notes
Loading notes...