The Classification Engine Explained

The Classification Engine

Klarvo's Classification Engine is the core logic that determines your AI system's risk level and applicable EU AI Act obligations.

How Classification Works

The engine processes your wizard answers through four sequential stages:

Stage 1: AI System Definition → Is this even an AI system?
     ↓
Stage 2: Prohibited Screening → Any red flags under Article 5?
     ↓
Stage 3: High-Risk Screening → Annex III category match?
     ↓
Stage 4: Transparency Check → Article 50 obligations?
     ↓
Final Classification + Obligation Mapping

Stage 1: AI System Definition

First, we determine if your system meets the EU AI Act definition of an "AI system":

Key criteria:

Possible outcomes:

Stage 2: Prohibited Practices (Article 5)

Critical safety check against eight prohibited AI practices:

Possible outcomes:

⛔ If any prohibited indicator is flagged, the system is classified as "Blocked" until cleared by legal review.

Stage 3: High-Risk Screening (Annex III)

Check against nine high-risk use case categories:

Possible outcomes:

Stage 4: Transparency Check (Article 50)

Evaluate transparency disclosure requirements:

Final Classification Output

Based on all stages, systems are classified:

Classification Confidence

Each classification includes a confidence level:

Audit Trail

Every classification decision is logged with: