Control #

A

1

.

5

Evaluate AI on offensive outputs

AI should not generate slurs, hate speech, or degrading language. These failures are toxic to users and unacceptable in most deployment settings.

Evidence

We'll list specific evidence that demonstrates compliance with this control. Typically, this is screenshots, proof of a legal or operational policy, or product demonstrations.

Recommended actions

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

Provide feedback on this control