EU AI Act duties depend on what the system does, its risk tier, and whether you are the provider or the deployer. So this is a decision workflow that establishes your actual obligations — not a checklist that asks everyone the same questions.
Almost every EU AI Act mistake starts there. The wizard walks a system through classification, establishes your role, and produces the obligation set that actually applies — recording the rationale, which is the thing a regulator asks about first.
Fund / Fix / Freeze triage across every AI initiative
| System | Risk tier | Role | Obligations |
|---|---|---|---|
| Credit scoring | High risk | Deployer | 12 duties |
| CV screening | High risk | Deployer | 12 duties |
| Support chatbot | Limited | Deployer | 3 duties |
| Doc summarisation | Minimal | Deployer | Voluntary |
Portfolio through to the specific legal obligation.
Every AI initiative in one view with triage matrix, model-risk distribution and maturity gauge.
Inventory with four risk tiers, status workflow and impact assessments.
Registry with training-data lineage, bias monitoring and deployment status.
Organisational readiness across eight domains, in full, quick or domain-specific modes.
AI-specific risk assessments with automated extraction and control mapping.
Guided classification, then the obligations that apply to your role and risk tier.
Fundamental-rights impact assessments for the systems that require them.
Every module shares one system of record, so evidence gathered in one place counts everywhere it is needed.
See the classification wizard run against a real system.