EU AI Act for public administration.
Public sector uses some of the most regulated AI categories: welfare eligibility, law enforcement risk scoring, immigration decisions, judicial support. Annex III designates many of these as high-risk. Matproof brings AI Act governance to public-sector procurement and operations.
Why this matters now
Several EU Member States are running public-sector AI systems now facing regulatory scrutiny. The AI Act's high-risk categories for public administration apply since August 2026 — government procurement is already demanding AI Act compliance attestations from vendors.
- Public procurement cycles make rapid AI Act adoption hard
- Cross-agency data sharing amplifies AI Act + GDPR obligations
- Citizen-facing decisions create explainability and recourse requirements
- Legacy public-sector AI (welfare benefit scoring, immigration triage) often pre-dates AI Act awareness
How Matproof covers EU AI Act for Public Sector
Annex III high-risk inventory
Public-sector high-risk categories: law enforcement (migration, border, prosecution support), administration of justice, access to essential private and public services (including welfare), education access, employment decisions. Map every AI system against these categories.
Fundamental rights impact assessment (FRIA)
Art. 27 FRIA is mandatory for public-sector deployers of high-risk AI. Matproof's template FRIA framework aligned with emerging guidance from EDPB and national authorities.
Human oversight and recourse
Citizens affected by high-risk AI decisions have explicit rights. Operational processes for human review, explanation, and appeal must be implemented and documented.
Public registration
Public-sector deployers of high-risk AI systems must register in the EU database (Art. 49 + 71). Matproof prepares the registration dossier from the AI inventory.
In scope
- Federal ministries and Bund agencies
- State authorities and social welfare administration
- Law enforcement and internal security agencies
- Immigration and asylum authorities
- Tax and revenue administration
- Public employment services and labor agencies
- Judiciary (supporting AI in case administration)
Frequently asked questions
Is a welfare-benefit eligibility AI system high-risk under the AI Act?+
Yes, explicitly — Annex III(5)(a) covers AI used for evaluating access to essential public services and benefits. Full high-risk obligations apply: risk management, data quality, human oversight, transparency to users, FRIA, registration in the EU database.
What is a Fundamental Rights Impact Assessment (FRIA)?+
Art. 27 requires public-sector deployers of high-risk AI to assess the system's impact on fundamental rights before deployment. Similar structure to a DPIA but broader — covers rights to non-discrimination, due process, human dignity, and procedural fairness. Must be documented, reviewed periodically, and available to supervisory authorities. Matproof provides a FRIA template and workflow.
How does AI Act apply to procurement of AI systems?+
Public-sector procurers must verify that AI systems they deploy meet AI Act obligations — Art. 26 deployer obligations plus the provider's Art. 9-15 obligations. Contracts should require compliance attestations, audit rights, and remediation obligations. Matproof's procurement templates include AI Act clauses.
Ready to start with EU AI Act?
30-minute demo tailored to Public Sector. We show you exactly how Matproof covers EU AI Act for your sector.