EU AI Act Readiness & Risk Classification

Tool and service

AIActSteps is the tool we use to classify AI applications against Article 5 (prohibited), Article 6 / Annex I & III (high-risk), Article 50 (transparency), and GPAI rules, mapping technical evidence against regulatory requirements. The risk analysis, architecture review, and governance design are the SKOPION Intelligence service. A tool replaces neither authority supervision nor formal third-party conformity assessment.

How it runs

1. 01System inventory & role identification: Cataloging AI models and distinguishing between provider, deployer, distributor, and importer responsibilities.
2. 02Risk classification: Evaluating prohibited practices (Art. 5), high-risk use cases (Annex I & III), and General-Purpose AI (GPAI) tiers (Art. 51+).
3. 03Governance & compliance audit: Auditing the Risk Management System (Art. 9), training data governance (Art. 10), logging (Art. 12), and human oversight (Art. 14).
4. 04Transparency & technical file: Reviewing Article 50 disclosure duties (AI content tagging, synthetic media) and Annex IV technical documentation.
5. 05Roadmap & remediation: Prioritised action plan aligned with statutory deadlines from 2025 through 2027.

What you receive

• A documented and reasoned classification of all assessed AI systems under Regulation (EU) 2024/1689.
• A structured gap assessment against Articles 9–15 (risk, data, logging, transparency, oversight, robustness).
• A technical verification report on Article 50 user-transparency and watermarking compliance.
• A phased implementation roadmap aligned with EU AI Office and national authority milestones.
• An executive and technical report in English, German, or Polish.

What this is not

• Not a conformity assessment issued by a Notified Body.
• Not formal legal counsel; legal interpretations are confirmed with qualified legal counsel.
• Not an official clearance or endorsement by the EU AI Office or national market surveillance bodies.
• Not a generic questionnaire; classification is grounded in verified system architecture and intended purpose.

Frequently Asked Questions on the EU AI Act

Which AI systems are covered by the EU AI Act (Regulation (EU) 2024/1689)?

The regulation distinguishes prohibited AI practices (Art. 5), high-risk AI systems (Art. 6 with Annex I and III), AI systems subject to transparency obligations such as chatbots and generative AI (Art. 50), and general-purpose AI models (GPAI, Art. 51+).

What do Article 14 (human oversight) and Article 15 (accuracy & cybersecurity) require?

Providers and deployers must ensure that high-stakes AI outputs can be effectively overseen or overridden by qualified staff (Art. 14) and that systems are technically resilient against prompt injection, data poisoning, and systematic bias (Art. 15).

How does SKOPION support organisations with EU AI Act readiness?

We inventory and classify your AI use cases using AIActSteps, evaluate technical safeguards (human-in-the-loop abstention, prompt-injection defense, audit logging), and deliver a prioritized compliance and engineering roadmap.