ai-inventory

To maintain an EU AI Act compliance inventory, you can capture each AI system's role and risk tier with visible reasoning, preventing the mixing of company-level and system-level obligations. The inventory supports listing, adding, editing, and classifying records.

EU AI Act classification covers roles like provider, deployer, importer, distributor, and product manufacturer, alongside risk tiers including prohibited, high-risk, limited, minimal, GPAI, and GPAI+systemic based on system characteristics.

To add a new AI system to the compliance registry, run the add command and provide the system name, owner, description, status, and EU nexus, which stores the record and offers to run the classification walkthrough next.

Yes, the AI system risk tiering process forces classification reasoning to be visible, marking role and tier basis for verification and routing into deeper analysis when needed rather than auto-deriving obligations from a static table.

You can track EU AI Act jurisdiction by capturing the EU nexus, which records whether the system is deployed in, offered to users in, or affects people in the EU/EEA to determine if the analysis applies.

The best way to classify an existing AI system is using the interactive classification walkthrough, which is driven by user-provided facts to determine the system's role and risk tier while marking the reasoning basis for verification.