Most investigations from riskbaserade urval lead to action

Sweden's Social Insurance Inspectorate found that about 74% of investigations started by Försäkringskassan's risk-based selections led to action in 2024, against 40% from public tips.

  • 2 an organization

    ISF found that about 74% of investigations started from risk-based selections led to action in 2024, against 40% of those started by public tips.

Depth runs from 1, a convenience or time saved, to 5, a life or a livelihood preserved.

Status
Reported benefit.
Evidence

foundstudiedreportedclaimed

An independent audit report supports the benefit.

Compared with

Control investigations started by tips from the public in 2024. Measured.

Without the system
40 procent
With it
cirka 74 procent
The system’s part
Not isolated. ISF compares investigations started from risk-based selections with those started by tips, which concern different cases. In 2024 targeted checks based on indicators of wrong payments produced 3,843 of the agency's risk-based impulses. Selection profiles that the control function develops produced 485. ISF does not say whether either kind learns from data. The data protection authority found that the agency had used a machine-learning risk profile for targeted checks of temporary parental benefit and had withdrawn it by September 2025.
Limits
Not yet checked.
Costs and harms
Not yet checked.
Later
Not yet checked.

Download data Cite Suggest a correction

Benefit type
Improved Task Accuracy
System
Försäkringskassan's risk-based selections (riskbaserade urval) for control investigations: targeted checks (riktade kontroller) and selection profiles (urvalsprofiler)
Made by
Försäkringskassan (Swedish Social Insurance Agency), including a unit of its control function that develops selection profiles
Used by
Försäkringskassan (Swedish Social Insurance Agency), control function
Where
Sweden
Stage
In routine use
Date
Kind
Aggregate result
Broad benefit
Analyzing Data or Information, Organizing, Planning, and Prioritizing Work
Collections
Scrutiny and findings
Last revised
31 October 2026

Sources

  1. Audit report, by authors with no tie to the system.

    Inspektionen för socialförsäkringen (ISF). Kontrollutredningar vid misstänkta bidragsbrott: En granskning av Försäkringskassans och Pensionsmyndighetens arbete vid misstanke om ersättning på felaktiga grunder. ISF Rapport 2025:14, published 17 December 2025. Section 3.3, printed pp. 67-68, and Bilaga 1, Figur B4.

    Under 2024 ledde cirka 74 procent av de kontrollutredningarna till åtgärd, jämfört med 40 procent av kontrollutredningarna som hade startats på grund av ett tips från allmänheten.

  2. Regulatory document, by authors with no tie to the system. Context or contrary evidence.

    Integritetsskyddsmyndigheten (IMY). Beslut efter tillsyn enligt dataskyddsförordningen – Försäkringskassan. IMY-2025-11828, 18 November 2025.

    IMY konstaterar att Försäkringskassan använt sig av en maskininlärningsbaserad riskprofil inom ramen för den aktuella handläggningen men att denna tagits ur bruk.

Details

AI system

Workflow

The selections analyzed patterns in the agency's registers and data systems to find cases with a high risk of wrong payments. Under the agency's rules, an impulse from a risk-based selection in principle always leads to a control investigation, without the investigator's own assessment.

Follow-up

Conditions for replication

The legal department must clear the data for each profile against discrimination and data protection law. Investigators must have time for risk-based cases. Interviewees said that the many external tips crowd these cases out.

Entry details

Sector

Social insurance control.

Purpose

Försäkringskassan built its risk-based selections to find cases with a high risk of wrong payments for control investigation.

Not yet checked: model, release, provider, role, configuration, inputs and outputs, autonomy, human oversight, AI contribution, beneficiaries, reach, distribution and access, duration, every measurement question, every money question, ownership and control, dependencies, cross-border connections, geopolitical implications, replication attempts, availability, institutional decision, grounds, action and result, success criterion, every evidence question, observed or intended, alternative assessments, related entries, other identifiers, category mappings.

Cite this record

Peter H. Vartanian. (2026). Most investigations from riskbaserade urval lead to action (AIBD-0919). Artificial Intelligence Benefits Database, edition of 31 October 2026. https://petervartanian.xyz/aibd/records/AIBD-0919.html (archived at https://doi.org/10.7910/DVN/AADSII, version 4.0)

BibTeX Data Field definitions