As part of EU AI Act implementation, Article 4 entered force on February 2, 2025, imposing a binding obligation on providers and deployers of AI systems to ensure “sufficient level of AI literacy” among staff and other persons handling AI systems on their behalf. Denmark implemented this through the Supplementary Provisions Act (2025), with enforcement responsibility distributed between the Agency for Digital Government and the Data Protection Agency.
The requirement applies across public and private sectors. Full enforcement mechanisms and supervision rules for the wider AI Act take effect August 2, 2026. Denmark’s approach emphasizes risk-based enforcement, public sector scaling, and ethical innovation.
Who it affects: Organizations deploying or providing AI systems; staff and contractors interacting with those systems; public sector agencies; private businesses and SMEs.
What is notably missing: No mandatory curriculum standard or proficiency definition for “sufficient AI literacy”; no public disclosure requirement for organizations’ training programmes; no worker right to paid training time; no protection against dismissal for raising AI literacy concerns; no mechanism to verify competency beyond self-certification by organizations; no ongoing compliance auditing by regulators until August 2026.