Regulatory AI conformity. With evidence.

EU AI Act, FINMA 08/2024, nDPA. We test your AI systems and deliver audit-ready reports.

Compliance Products

Entry
AI Risk Classification
EU AI Act risk classification for your AI system. Structured assessment using our structured AI Act risk classification system with Swiss regulatory overlay (nDPA + FINMA). Determines risk level, applicable articles, compliance obligations, and timeline.
from CHF 3,000 1 week
Recurring: Annual re-classification
Next: Compliance Report
Premium
FINMA Model Validation
Purpose-built evaluation against FINMA Guidance 08/2024. First benchmark-based FINMA AI validation in Switzerland.
  • 5 categories, comprehensive FINMA-specific evaluation
  • Board-ready executive summary
  • Risk heat map with remediation roadmap
from CHF 15,000 2–4 weeks
Recurring: Annual re-validation (from CHF 5,000)
Start with our free Compliance Scorecard or download the Quarterly Report →
Enforce compliant model routing automatically. The AI Model Router ensures only compliant models handle high-risk tasks. Routing rules derived from your compliance assessment. Learn more →

Key deadlines and requirements

The EU AI Act requires technical compliance evidence for high-risk AI systems. Annex III systems face a December 2027 deadline; Annex I systems face August 2028. Almost no Notified Bodies have been designated yet.

FINMA Guidance 08/2024 requires traceable model validation for AI systems in financial services. FINMA Guidance 08/2024 and the federal AI law consultation (by end 2026) will tighten requirements further.

The revised Swiss Data Protection Act (nDPA) adds data sovereignty requirements that intersect with AI governance.

What you receive

  • Per-article compliance scores (Art. 9–15 EU AI Act)
  • Traffic-light risk dashboard
  • Gap analysis with remediation priorities ranked by regulatory risk
  • FINMA 08/2024 alignment report (for Premium engagements)
  • Methodology documentation for independent verification
  • Board-ready executive summary

Air-gapped evaluation for regulated institutions

For FINMA-regulated institutions and high-security environments where data cannot leave the premises: we bring the evaluation to you. Our dedicated hardware runs the complete evaluation pipeline on-site, fully air-gapped. No data leaves your network, no cloud dependencies, no third-party access.

Available as part of any compliance engagement. See all data handoff modes →

Every compliance assessment reveals performance implications. How does your model selection affect compliance scores? See our Performance evaluations →

Schedule a scoping call

Start with a Risk Classification or go directly to a full EU AI Act assessment. First step is always a scoping call. No preparation needed.

The EU AI Act high-risk deadline is December 2027: 20 months away. Compliance projects take 6–12 months. Our approach is grounded in peer-reviewed methodology (ArXiv, 2026). Start now.

contact@ai-helvetic.ch