NovaCopilot AI Act Classification API
Which risk class does this AI system fall into, and what is required before launch?
Classifies an AI system under the EU AI Act and returns the obligations, conformity steps and transparency duties that apply to the provider or deployer.
AI Act · GDPR · contract preview
Overview
NovaCopilot AI Act Classification API is part of the NovaCopilot Trust Layer. Input is evaluated against the NovaCopilot Trust Graph by the NovaCopilot Decision Engine and returned as structured JSON with sources. Status: contract preview.
Input
- use_case
- The intended purpose of the AI system
- role
- provider or deployer
- country
- ISO country code
- generates_content
- Whether the system produces synthetic content
Output
- decision
- allowed, allowed_with_controls or not_allowed
- risk
- low, medium or high
- confidence
- Number between 0 and 1
- eligibility_status
- NovaCopilot Eligibility Graph: allowed, restricted, high_risk or not_allowed
- restrictions
- What may not be done for this input
- alternative_paths
- Lower-risk ways to reach the same goal
- applicable_regulations
- Regulations that apply to the input
- required_controls
- Controls required before going ahead
- summary
- Plain-language explanation
- reasoning_summary
- Plain-language explanation of the decision path
- affected_groups
- Groups affected, for example minors or adults
- sources
- NovaCopilot Trust Graph objects behind the decision: type, regulation, article, jurisdiction, effective date and official source link
- relevant_enforcement
- NovaCopilot Enforcement Intelligence: earlier supervisory decisions, fines, court decisions and guidance on similar conduct, with authority, date and source
- regulatory_signals
- Regulatory trends that affect the input
Example Request
POST /api/public/v1/agents/ai-act-classification
{
"use_case": "recruitment_screening",
"role": "deployer",
"country": "DE",
"generates_content": false
}Example Response
{
"decision": "allowed_with_controls",
"risk": "high",
"confidence": 0.86,
"applicable_regulations": [
"AI Act",
"GDPR"
],
"required_controls": [
"Human oversight",
"Risk management system",
"Inform affected workers",
"Fundamental rights impact assessment where required"
],
"summary": "Employment screening is listed as a high-risk use case."
}Use Cases
- Launch reviews for new AI features
- Vendor due diligence
- AI inventory and governance programmes
OpenAPI Schema
Machine-readable specification: /api/public/v1/openapi.json
JSON Contract
The response format is stable. Fields may be added, existing fields are not renamed or removed. Agent contract: /api/public/v1/agents/ai-act-classification
The registry is large. The first answer can take a few seconds to load.