Industry Guides

EU AI Act Compliance by Industry

The EU AI Act applies differently depending on what your AI system does and who it affects. These guides translate the regulation's Annex III categories into concrete obligations for specific industries — so you know exactly what applies to your product.

Annex III, Category 4

HR Technology

CV screening, interview assessment, performance evaluation, and employee monitoring tools are explicitly listed as high-risk AI.

ATS and CV screeningInterview scoringPerformance managementTask allocation AI
Annex III, Category 5

Fintech & Credit

AI used for creditworthiness assessment, insurance risk scoring, and access to essential financial services is high-risk under the Act.

Credit scoringLoan decisioningInsurance pricingFraud detection
Annex I / Annex III

Healthcare AI

Healthcare AI splits across two regimes: medical device AI (Annex I, 2027 deadline) and administrative health AI affecting patient access (Annex III).

Clinical decision supportPatient triageDiagnostic AICare pathway tools
Annex III, Category 3

Edtech

AI systems that determine access to education, assess learners, or monitor student behaviour are explicitly high-risk under Annex III.

Admissions AIAutomated gradingProctoring softwareLearning analytics
Annex III, Category 8

Legal Technology

AI deployed in courts and tribunals to assist judicial decisions is explicitly high-risk. Legal chatbots and AI-generated content face Article 50 transparency obligations.

Court decision supportLegal chatbotsAI-generated contractsLegal research tools
Annex III, Category 5

Insurance AI

Insurance underwriting, claims decisioning, and risk scoring AI falls under Annex III Category 5 — the same category as credit scoring.

Underwriting AIClaims decisioningFraud detectionTelematics pricing
Annex III, Category 5

PropTech & Real Estate

Tenant screening, rental application scoring, and mortgage-linked property valuation AI fall under Annex III Category 5. Housing is a fundamental right — expect close scrutiny.

Tenant screeningAutomated valuationsRental scoringApplication ranking
Article 50 — Transparency

Customer Service AI

Customer service chatbots are typically not high-risk, but every chatbot that interacts with users must disclose its AI nature before the conversation starts under Article 50(1).

Customer chatbotsAI voice assistantsAI-generated email responsesSentiment analysis

Not sure which category applies to you?

Nytivo's risk classification wizard walks you through the Annex III categories and identifies your obligations in 5 minutes.

Check your risk tier free