Artificial intelligence (AI) and automated decision-making systems are being introduced in social administration across Europe. Decisions on social benefits are often based on algorithmic risk assessment and the creation of citizen profiles.
However, automation in the social sector follows technical logic and neglects socio-political consequences, claims ITA researcher Doris Allhutter in her new ITA dossier ‘Automation in the welfare state’. In order to be able hold state institutions accountable of their duties, transparency and institutional strategies would be essential, stresses Allhutter.
Which data is used? What assumptions do algorithms make about the social situation of citizens, and how are automated decisions generated? All these questions have huge consequences for citizens and are therefore important social policy issues, says Allhutter. Rather than just applying the latest technology, Allhutter recommends to first clearly communicate on what basis people are assessed by authorities: "To strengthen democracy, state institutions must act transparently in the design and implementation of social policy."
Links
ITA-Dossier ‘Automation in the welfare state’
(Findings and recommendations based on the paper ‘Public value in the making of automated and datafied welfare futures’, published in the Internet Policy Review).