EU AI Act · REST API

EU AI Act Compliance Scanner

Point the API at any GitHub repository or describe your AI system.
Get a risk score, affected articles, and actionable fixes — in under 100ms.

Try the API → See pricing →

No sign-up for a quick test  ·  OpenAPI spec available  ·  50 free scans/month

One call. Full compliance picture.

Pass a GitHub URL or inline code. The API returns risk score, category, and per-article recommendations.

Request

curl -s -X POST https://arkforge.fr/euaiact/v1/scan \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://github.com/openai/openai-python",
    "sector": "finance"
  }'

Response

{
  "scan_id": "38c47a18-fb24-4bc0-8498-ab297093832a",
  "risk_score": 33.3,
  "risk_category": "limited",
  "frameworks_detected": ["openai", "anthropic"],
  "files_scanned": 1208,
  "compliance_score": "2/3",
  "recommendations": [{
    "check": "content_marking",
    "status": "FAIL",
    "eu_article": "Art. 52(3)",
    "effort": "low",
    "how": ["Add metadata or visible label to AI-generated outputs", ...]
  }]
}

What the API checks

Each scan validates your system against the key provisions of the EU AI Act (Regulation 2024/1689).

Art. 6 — Risk classification
Identifies your system's risk tier: minimal, limited, high, or unacceptable.
Art. 13 — Transparency
Checks that your AI system is documented and explainable to users.
Art. 14 — Human oversight
Verifies that human control mechanisms are present.
Art. 52(3) — Content marking
Ensures AI-generated outputs are labeled as required.
Art. 9 — Risk management
Detects missing risk management systems and logging.
Art. 17 — Quality management
Checks for quality assurance processes and documentation.

Risk categories

Limited

Risk score 0–40. Minor transparency obligations. Low compliance effort.

High Risk

Risk score 40–80. Full Art. 9-17 requirements apply. Audit trail mandatory.

Unacceptable

Risk score 80–100. Prohibited practices detected. Immediate action required.

Pricing

Available on arkforge.fr/euaiact/docs and on RapidAPI.

Free
€0 /month
  • ✓ 50 scans/month
  • ✓ GitHub URL scan
  • ✓ Full recommendations
  • ✗ No API key required for basic use
Try now →
Pro
€19 /month
  • ✓ 500 scans/month
  • ✓ Async batch scanning
  • ✓ Usage tracking endpoint
  • ✓ Priority response
Get Pro →
Ultra
€49 /month
  • ✓ 2 000 scans/month
  • ✓ All Pro features
  • ✓ Sector-specific analysis
  • ✓ SLA <100ms p99
Get Ultra →

API endpoints

Method Path Description
GET /health Service status
POST /v1/scan Sync scan (URL or code)
GET /v1/scan/{id} Retrieve previous scan
POST /v1/scan/async Async scan (large repos)
GET /v1/job/{id} Async job status
GET /v1/usage Monthly scan usage
Open full API docs →

FAQ

Does the EU AI Act apply to my product?

If your product uses AI and is placed on the EU market or used in the EU, yes. The regulation applies to providers, deployers, and importers regardless of where they are incorporated. The scanner tells you your risk tier in seconds.

When does the EU AI Act come into force?

August 2024 (entry into force). Prohibited practices: February 2025. High-risk system obligations: August 2026. GPAI model rules: August 2025. Start compliance work now.

What does the API actually scan?

For GitHub repos: all source files are fetched and analyzed for AI framework usage, data handling patterns, and documentation gaps. For inline code: direct analysis. Sector context adjusts the risk scoring.

Is there a limit without authentication?

The free tier allows 50 scans/month with no API key. For higher volumes, contact us for a Pro or Ultra plan.