Module 4: AI Regulation · BoK IV.C
High risk - where most regulation lives
Last reviewed:
High risk / high-impact AI significantly affects rights, safety or access to essential services. It is allowed but under strict obligations, and the majority of AI regulation applies here - including healthcare, credit, housing and employment decisions Colorado labels consequential decisions.
Systems that significantly affect rights, safety or access to essential services. Allowed, but under strict obligations. The majority of AI regulation applies here.
| Category | Jurisdictional anchors |
|---|---|
| 🏗️ Critical infrastructure & safety → transport, energy, medical devices | EU Annex II; South Korea requires risk management and documentation for high-impact sectors |
| 🏥 Healthcare, credit, housing, employment decisions → access to jobs, loans, housing, care | EU Annex III; Colorado SB 24-205 calls these consequential decisions |
| 🎓 Education & testing → scoring exams, assessing students, allocating opportunities | EU Annex III; South Korea's high-impact designation includes education |
| 🚔 Law enforcement & justice → predictive analytics, biometric ID in investigations, risk scoring | EU Annex III; multiple jurisdictions add oversight or restrictions |
| 🤖 Generative AI with systemic risks → foundation and general-purpose models with wide impact | EU introduces the systemic risk category; China requires safety reviews and security filings; California mandates training data transparency |
| Sensitive biometric uses → facial, voice or emotional analysis risking unfair classification or monitoring | EU restricts; China and US state laws impose disclosure or consent duties |
Key terms - quick answers
What is “Consequential decisions”?
Colorado SB 24-205 term for high-risk decisions on jobs, loans, housing and care, triggering an AIA.
Keep going - free
Every study note on this site is free. So are these: