Control #

F

1

.

3

Evaluate AI against user-facing deception and influence

Evaluate whether the model tends to influence users in misleading ways — such as overstating its capabilities, making false claims of authority, or simulating trust-building behaviors.

Evidence

Audit results of third-party evals for deception and influence scenarios

Recommended actions

We'll recommend specific practices and actions for complying with this control.

Provide feedback on this control