EU AI Act Risk Classifier

Q0: Agentic AI? (Makes decisions alone)

• YES: TREAT AS HIGH STAKE- LAWYER NEEDED
• NO: 10 questions → Tier output.

Q0

• [ ] YES
• [ ] NO

Questions (answer and keep notes)

1. Does it hire people? [Yes/No]
2. Uses faces? [Yes/No]
3. Does it impact access to education? [Yes/No]
4. Does it impact access to financial services (credit/insurance)? [Yes/No]
5. Does it support law enforcement or public authority decisions? [Yes/No]
6. Is it used in workplace management (monitoring/hiring/firing)? [Yes/No]
7. Is it used in healthcare/medical decision support? [Yes/No]
8. Does it perform biometric identification or categorization? [Yes/No]
9. Is it used in critical infrastructure (energy/water/transport)? [Yes/No]
10. Does it rank or decide people for high-stakes outcomes? [Yes/No]

Output

• Result: [Low / Limited / High]
• Notes: [why]