AI System Inventory
Register every model/provider/use case, purpose, user group and data classification.
platform ARCHITECTUREAI Governance
Inventory first. Classify the use case. Then apply the right controls.
This platform page translates the current EU AI Act timeline into operational architecture for the platform. It is not legal advice and should be revalidated against official EU guidance before deployment decisions.
Use-Case Classification
Assess whether a feature is prohibited, transparency-limited, general-purpose or potentially high-risk.
HUMAN / LEGAL REVIEWAI Literacy
Maintain role-appropriate training and guidance for staff who use or supervise AI systems.
POLICY + TRAININGLogging & Evidence
Preserve model version, evaluation, prompts/policies, approvals and significant outputs where appropriate.
platform DATA MODELHuman Oversight
Define who may override, stop, review or approve consequential model-assisted decisions.
REQUIRED DESIGNMonitoring
Track incidents, drift, harmful failure patterns and changes in provider/model behaviour.
OBSERVABILITYRegulatory snapshot date: 11 September 2026. High-risk classification depends on the exact function, context and role of the organisation. Obtain appropriate legal/compliance advice for production deployments.