XArticle index · Reg. (EU) 2024/1689
Compliance features filed against every article.
Treda maps its feature set directly to EU AI Act articles and annexes, so you can demonstrate compliance with confidence. Each feature traces back to specific regulatory requirements.
IVRisk classification
Four tiers, four dockets.
Article 6 defines four risk tiers. Treda helps you classify and manage obligations for each.
Unacceptable
IArticle 5
AI practices banned outright. Includes social scoring by governments, real-time biometric identification in public spaces (with limited law enforcement exceptions), and manipulation of vulnerable groups.
- Subliminal manipulation techniques
- Exploitation of age, disability, or social situation
- Social scoring by public authorities
- Untargeted facial image scraping for recognition databases
High Risk
IIArticles 6–7, Annex III
AI systems used in critical areas listed in Annex III. Subject to the full set of obligations under Chapters III–IV: risk management, data governance, technical documentation, transparency, human oversight, and accuracy and robustness.
- Credit scoring and loan eligibility decisions
- AI in medical devices (diagnosis, treatment recommendations)
- Recruitment and worker management systems
- Law enforcement risk assessment tools
- Migration and border control systems
- Biometric identification (with authorization)
Limited Risk
IIIArticle 50
AI systems with specific transparency obligations. Users must be informed they are interacting with an AI system or viewing AI-generated content.
- Chatbots and virtual assistants
- Emotion recognition systems
- Deepfake generators and AI-generated text, image, and video
- AI-powered recommendation engines (at member state request)
Minimal Risk
IVArticle 6(3)
Most AI systems fall here. No mandatory obligations, but voluntary codes of conduct are encouraged. Examples include AI-enabled video games, spam filters, and inventory management systems.
- AI-enabled video games and entertainment
- Spam filtering and email classification
- Inventory management and logistics optimization
- Weather forecasting models
FCore filings
The dossier, section by section.
Each feature corresponds to specific EU AI Act articles and obligations.
AI System Registration & Inventory
Maintain a live register of every AI system your organization deploys, including intended purpose, provider identity, and risk classification per Article 26. The register maps directly to the EU Database requirements for high-risk systems.
- Automated discovery scans for shadow AI across cloud providers
- Per-system lifecycle tracking: development, testing, deployment, withdrawal
- Mapping to EU Database registration fields (Article 26(5))
- Version-controlled model cards with training data provenance
Risk Classification Engine
Classify AI systems against the four-tier risk framework. The engine cross-references your system's intended purpose and use case against Annex III prohibited and high-risk categories.
- Guided questionnaire based on Articles 6(1)–6(3) classification criteria
- Annex III sector-by-sector mapping (employment, education, law enforcement, etc.)
- Automatic escalation for biometric identification and remote biometric systems
- Change-in-purpose re-classification alerts when use cases evolve
Conformity Assessment Workflows
Orchestrate the full conformity assessment process for high-risk AI systems. Build and maintain the technical documentation file, apply Annex IV standards, and prepare for CE marking.
- Checklist-driven workflow aligned with Article 43 conformity assessment procedures
- Internal assessment support for non-Annex III systems (Article 43(1))
- Third-party notified body coordination for Annex III systems requiring external review
- CE marking documentation generation with EU Declaration of Conformity (Article 47)
Technical Documentation Generator
Auto-generate and maintain technical documentation meeting Annex IV requirements. Living documents that update as your AI system evolves throughout its lifecycle.
- Annex IV section coverage: general description, development process, monitoring
- Training data governance documentation (data sources, representativeness, gaps)
- Model architecture specifications and performance metrics per Article 15
- Automated impact assessments for fundamental rights (Article 27)
Incident Reporting & Tracking
Track, report, and resolve serious incidents per Article 62 obligations. Configure automated escalation rules, notify market surveillance authorities, and maintain a post-market monitoring system.
- Incident severity classification (serious vs. minor per Article 62(1))
- Automated notification to market surveillance authorities within required timeframes
- Root cause analysis workflows with corrective action tracking
- Post-market monitoring plan management per Article 72
Human Oversight Provisions
Document and verify that high-risk AI systems incorporate effective human oversight per Article 14. Ensure operators can interpret outputs, override decisions, and intervene when necessary.
- Oversight mechanism design documentation (human-in-the-loop, human-on-the-loop)
- Operator training record management and competency verification
- Override and intervention logging with full audit trail
- Regular oversight effectiveness reviews with KPI dashboards
Transparency & Disclosure
Manage transparency obligations under Article 50 for limited-risk systems and Article 13 for high-risk systems. Ensure users are properly informed about AI interactions and decision-making.
- Automated disclosure generation for chatbot and deepfake interactions
- User notification templates per Article 50(2) transparency requirements
- AI-generated content watermarking and provenance metadata
- High-risk system output explanations per Article 13(3)(b)
Continuous Monitoring & Alerting
Real-time monitoring dashboards for AI system performance, drift detection, and compliance status. Alert on anomalies that may trigger re-assessment or incident reporting requirements.
- Model performance drift detection with configurable thresholds
- Compliance deadline tracker (enforcement dates vary by risk tier)
- Regulatory change monitoring for EU AI Act amendments and guidance
- Automated evidence collection for audit readiness
Ready to classify your AI systems?
Start with our guided risk classification questionnaire and get a compliance roadmap tailored to your AI portfolio.