System summary
Name the AI feature, user group, deployment region, vendor, input data, output type, and business owner. A short summary prevents a guide result from becoming detached from the actual product behavior.
Startup checklist
Founders do not need to turn every AI feature into a legal project. They do need a reliable way to spot consequential use cases before launch.
If your AI changes what a person can access, buy, receive, qualify for, or appeal, treat it as a specialist-review use case.
Last reviewed: 2026-07-03.
Implementation example
Use this example structure when saving a result from this guide into an internal AI compliance folder. The goal is to make the output specific enough that another teammate can understand the decision months later.
Name the AI feature, user group, deployment region, vendor, input data, output type, and business owner. A short summary prevents a guide result from becoming detached from the actual product behavior.
List whether the system affects employment, education, credit, healthcare, biometric identification, public benefits, safety, or another high-impact area. If the page is used for transparency wording, record where the notice will appear and what users can do after seeing it.
Set a clear trigger for review: model change, new vendor, new data category, new market, product launch, user complaint, policy update, or a change from internal-only use to customer-facing use.