Udbetaling Danmark's 60-Model Fraud System — Denmark's AI Welfare State Faces an Ombudsman Inquiry

Udbetaling Danmark (UDK), the agency that has centralised most Danish welfare payments since 2012 in partnership with pension fund ATP, uses more than 60 algorithmic and machine-learning models — built by IT contractor NNIT — to flag applicants and claimants for suspected benefits fraud investigations.
Amnesty International's November 2024 report 'Coded Injustice' obtained partial documentation on only 4 of the 60 models, but found that named models such as 'Really Single' (which flags 'atypical' cohabitation or living patterns) and 'Model Abroad' (which flags 'medium to high' ties to non-EEA countries) systematically increase scrutiny of people with disabilities, low-income individuals, migrants and refugees, and people in non-traditional households.
Amnesty argues the fraud-control system likely falls under the EU AI Act's 'social scoring' prohibition, which took effect on 1 August 2024. The Danish Parliament debated the findings in spring 2024, but the governing coalition declined to suspend the models, promising only 'ongoing internal monitoring'; the Parliamentary Ombudsman opened a formal inquiry in June 2024 into flagship models including the 'Gladsaxe Model'.
The case shows that even one of the world's most digitised welfare states can run a large, decade-old portfolio of opaque automated risk-scoring models — with only partial external visibility — for years before an independent body formally examines it.

Read the full analysis: https://www.computerweekly.com/news/366615893/Denmarks-AI-powered-welfare-system-fuels-mass-surveillance