← AI at Invent / How we govern AI
Self-assessment

Score your readiness before you engage.

Run this against your product before Gate 0. It mirrors the mandatory standards across all four domains, tells you where you stand, and shows you exactly what to close before a review. Nothing is stored or sent. It is your working view.

Step 1: Set your risk class

If you're not sure, pick the highest that could apply. The checklist adjusts what counts as mandatory.

Responsible & Safe AI

Safe, fair, transparent, accountable, before accurate.

Technical Architecture

Secure, scalable, evaluable, affordable to run.

Product & UX

The right problem, in an interface people trust.

Delivery & Quality

Testable, maintainable, safe to hand over.