Tools based on artificial intelligence are increasingly used in the implementation of social policies. On the positive side, some of these tools promise greater effectiveness and efficiency in the delivery of policies such as unemployment insurance or social assistance. Examples of such instruments include profiling tools that are used to segment jobseekers according to their needs in terms of re-employment services, or tools that automatize eligibility decisions on the basis of available administrative data.
On a more negative side, some of these tools have been criticized in terms of lack of accountability, risk to stigmatize disadvantaged populations and lack of transparency. In this track will first be presented the current state of affairs in AI and social policy. We will then discuss selected examples and assess their potentials and risks.