EU AI Act Non-High-Risk Compliance Roadmap

•
869
•

Guides the legal documentation and compliance workflow for AI systems classified as non-high-risk under the EU AI Act.

This skill provides a structured roadmap for developers and legal teams when an AI system has been classified as non-high-risk under the EU AI Act (KI-VO). It facilitates the creation of a 'negative diagnosis' documentation package, ensuring that the exclusion from high-risk categories (Annex I & III) or the application of exemptions (Art. 6 Abs. 3) is legally defensible. Beyond classification, it manages remaining obligations including transparency duties for chatbots and deepfakes (Art. 50), GPAI requirements, and AI literacy (Art. 4). Designed specifically for the German legal market, it references local databases and provides clear deadlines for implementation to avoid significant regulatory fines.

主な機能

01869 GitHub stars
02Checklist for prohibited AI practices and Art. 4 AI literacy
03Verification of transparency duties under Art. 50 KI-VO
04Structured 'Negative-Diagnosis' documentation for regulatory audits
05Integration with German legal databases like openJur and dejure.org
06Deadline tracking for EU AI Act phased implementation

ユースケース

01Ensuring transparency compliance for AI-generated content or customer-facing bots
02Preparing internal compliance packages for German companies utilizing GPAI models
03Documenting why an AI tool does not trigger high-risk requirements for market surveillance