Beyond the obligation for high risk systems to conduct a data protection impact assessment, the value added of impact assessments for IA models would rely on the risks covered, the economic dimension of such assessment and its interest in terms of IA governance. The panel will discuss the content and benefits of user-centric impact assessments for IA across the value chain, in particular:
- What are the risks to be covered by an AI impact assessment? What would be its objectives from a user-centric point of view?
- How can economic analysis be used in order to assess the impact of AI solutions and their regulation?
- How can the costs and benefits of applications of AI be effectively weighed against fundamental rights and safety in such an approach?
- At what stage(s) should an impact assessment of an algorithmic solution be best undertaken and by which entity(ies)?
