ai-inventory

To maintain an EU AI Act system registry, you need to capture each AI system's role and risk tier individually to scope obligations correctly. This approach prevents compliance errors caused by treating role and tier as company-wide rather than system-by-system.

Per-system risk classification records an AI system's role (provider, deployer, distributor) and risk tier (prohibited, high-risk, limited, minimal, GPAI) individually. This system-by-system evaluation ensures obligations are accurately scoped for each entry in your inventory.

You can add a new AI system to your compliance inventory workflow by triggering the add command. The system guides you through capturing required role and risk tier data, generating classification reasoning fields marked for attorney verification.

EU AI Act compliance tracking requires lawyer verification for role and tier reasoning fields. The workflow produces explicit classification tags for legal review and routes deeper obligation analysis to a dedicated conversation workflow instead of relying on hardcoded tables.

Yes, you can use a YAML file to track AI system roles and risk tiers. The inventory persists records to an ai-systems.yaml file, storing per-system role and tier classifications required for obligations scoping and review management.

You should avoid company-wide AI risk tier assignments because roles and tiers vary per system under the EU AI Act. Assigning a single tier across an organization causes compliance errors and fails to scope obligations correctly for individual AI systems.