Compliance/EU AI Act/Govtech & Public Sector
EU AI Act compliance for Govtech & Public Sector
AI used by (or on behalf of) public authorities for benefits eligibility (§5(a)) or emergency-call dispatch (§5(d)). Both high-risk. Art. 14 human oversight is operative — public-administration AI doesn't get to be unsupervised.
What applies
Articles + provisions of Regulation (EU) 2024/1689 — Artificial Intelligence Act most directly engaged by Govtech & Public Sector deployments:
- Annex III §5(a)
- Annex III §5(d)
- Art. 12
- Art. 14
What this looks like in Govtech & Public Sector
AI used by (or on behalf of) public authorities for benefits eligibility (§5(a)) or emergency-call dispatch (§5(d)). Both high-risk. Art. 14 human oversight is operative — public-administration AI doesn't get to be unsupervised.
Flagship examplePost-SyRI Dutch transparency expectations: per-decision evidence is the substrate of meaningful explanation.
Where Promethean's evidence layer fits
The substrate emits specific evidence kinds that map to EU AI Act obligations. Most directly relevant for Govtech & Public Sector:
For the full sector view including LLM-feature catalogue, tier recommendation, and reviewer-gate examples, see the Govtech & Public Sector sector page.
Adjacent cells
Same regulation, other sectors
Other regulations for Govtech & Public Sector