Artificial intelligence and algorithmic systems are increasingly used to help make decisions about public service provision and state benefits for citizens. Delegating decision-making to machines raises ethical and social concerns and important questions about responsibility, accountability, transparency, and the quality of such decision-making (see, for example, Allhutter et al. 2020). Countries also have different policy contexts where these systems are situated, which can impact the attitudes towards AI-supported decision-making and understanding of fairness in particular contexts. By using a “serious games” approach, we aimed to understand how a selected automated assessment system might be reviewed and improved with real-world stakeholders to incorporate concerns of fairness into it. For this purpose, we conducted half-day-long workshops with master students (n = 18) from different social science-focused curriculums who, by taking the roles of consultants or clients in an unemployment insurance fund, played out the various possible scenarios for clients when an algorithm is used to assess the needs of the clients. The chapter analyzes the workshop results with master students as real-life stakeholders inspired by the serious games approach and discusses further opportunities for better integration of AI-based tools into social service provision in Estonia.