Overlay GEN
Generative AI
Applies when the organisation uses AI that generates text, images, audio, video or code.
If this applies to your organisation, these 6 criteria are added to the controls below. If it does not, they are left out: they are neither met nor failed. Every other criterion applies to every organisation.
Review and disclosure of AI-generated content
- GOV-10.1 Approved rules set how AI-generated content released outside the organisation, including synthetic media, is reviewed and disclosed.
- GOV-10.2 Approved rules set how AI-generated content used inside the organisation to inform decisions or form official records is reviewed and disclosed.
- GOV-10.3 The rules require disclosure at least to the extent the law requires, made clearly when a person first meets the content.
AI risk assessment method
- RSK-02.7 The method covers the recognised generative AI risk categories.
Testing and evaluation before release
- TEC-06.5 Testing of generative AI covers harmful output.
Security testing of AI before release
- SEC-06.4 Testing of generative AI includes attempts to bypass its safeguards.