Is Your AI System High-Risk?
The EU AI Act's Annex III defines which AI systems face the strictest compliance requirements. Find out if yours qualifies.
Why Classification Matters
High-risk AI systems must meet all requirements in Articles 8-15 of the EU AI Act, including risk management, human oversight, and technical documentation. Non-compliance can result in fines up to €15 million or 3% of global revenue.
High-Risk AI Categories (Annex III)
The EU AI Act defines 8 categories of high-risk AI systems:
1 Biometrics
- Remote biometric identification systems
- Biometric categorization based on sensitive attributes
- Emotion recognition in workplaces and schools
2 Critical Infrastructure
- AI managing road traffic and vehicle safety
- Water, gas, heating, electricity supply management
- Digital infrastructure operations
3 Education & Training
- AI determining access to educational institutions
- Student assessment and evaluation systems
- Learning behavior monitoring
- Proctoring systems for exams
4 Employment & Workers
- CV screening and recruitment AI
- Job advertisement targeting
- Performance evaluation systems
- Promotion and termination decisions
- Task allocation based on behavior
5 Essential Services Access
- Credit scoring and creditworthiness assessment
- Insurance pricing and risk assessment
- Emergency services dispatch prioritization
- Public benefit eligibility evaluation
6 Law Enforcement
- Polygraphs and emotion detection
- Evidence reliability assessment
- Crime prediction (profiling)
- Criminal investigation analysis
7 Migration & Border Control
- Polygraphs for asylum applications
- Risk assessment for migration
- Visa application processing
- Identification verification
8 Justice & Democracy
- AI assisting judicial decisions
- Legal research and interpretation
- Alternative dispute resolution
- Election and voting systems
How to Determine If Your AI Is High-Risk
Check Annex III Categories
Does your AI system fall into any of the 8 categories listed above?
Check Safety Components (Annex I)
Is your AI a safety component of a product covered by EU harmonization legislation (machinery, toys, medical devices, etc.)?
Assess Material Impact
Does your AI significantly affect health, safety, or fundamental rights?
Consider Exceptions
Is your AI purely assistive, preparatory, or does a human make the final decision with full understanding?
High-Risk AI Requirements
If classified as high-risk, your AI system must comply with:
Article 9: Risk Management
Establish a continuous risk management system throughout the AI lifecycle.
Article 10: Data Governance
Ensure training data is relevant, representative, and properly governed.
Article 11: Documentation
Maintain comprehensive technical documentation before market placement.
Article 12: Record-Keeping
Implement automatic logging for traceability and audit purposes.
Article 13: Transparency
Design for sufficient transparency so users can interpret outputs.
Article 14: Human Oversight
Enable effective human oversight with ability to override AI decisions.
Article 15: Accuracy & Security
Ensure appropriate accuracy, robustness, and cybersecurity.
Conformity Assessment
Complete conformity assessment before placing on market.
Common High-Risk Use Cases by Industry
Financial Services
- Credit scoring algorithms
- Fraud detection affecting customer access
- Insurance underwriting AI
- Loan approval systems
Human Resources
- Resume screening tools
- Interview analysis systems
- Performance monitoring AI
- Workforce management algorithms
Healthcare
- Diagnostic AI systems
- Treatment recommendation engines
- Patient triage algorithms
- Medical imaging analysis
Not Sure About Your Classification?
Our free audit tool analyzes your AI system and determines its risk classification under the EU AI Act. Get instant results and compliance recommendations.
Check Your Classification