Försäkringskassan's discriminatory welfare-fraud AI (Sweden)

Försäkringskassan (the Swedish Social Insurance Agency) deployed a machine learning algorithm from at least 2013 to assign risk scores to social benefit applicants — primarily those claiming temporary child support — in order to prioritise cases for fraud investigation. The system operated without public awareness and was described internally as the agency's 'best-kept secret.'

In November 2024, Lighthouse Reports and Svenska Dagbladet published the investigation 'Sweden's Suspicion Machine,' revealing that the algorithm disproportionately flagged women, people with migrant backgrounds, low-income earners, and those without university degrees. Testing against six standard statistical fairness metrics (including demographic parity, predictive parity, and false-positive rates) confirmed systemic bias. Sweden's Inspectorate for Social Insurance (ISF) had reached similar conclusions in a 2018 supervisory report, finding the algorithm 'does not meet equal treatment' — but Försäkringskassan rejected those findings and continued operating the system.

Swedens' Integrity Committee had raised concerns as early as 2016 about extreme risks to citizens' personal integrity. The agency rejected virtually all freedom-of-information requests about the system over multiple years. Amnesty International called on Swedish authorities to discontinue the AI systems in November 2024.

Sweden's Data Protection Authority (IMY) launched an investigation following the Lighthouse Reports revelations; by June 2025, Försäkringskassan suspended the model. The agency subsequently conceded 'shortcomings in transparency and risk management' and added temporary manual oversight atop automated scoring.

This case parallels the Netherlands' SyRI and is a cautionary example of how ML risk-scoring in social security can embed structural inequality when deployed without adequate fairness auditing, meaningful redress mechanisms, or genuine regulatory scrutiny. It illustrates the particular accountability risks of opaque automated decision-making in welfare administration.

Read the full analysis: https://www.lighthousereports.com/investigation/swedens-suspicion-machine/