Concluding remarks
Open access

AI is increasingly present in the citizen-government relation today. It mediates many interactions between civil society and those in power. In this context, AI is mainly in the hands of governments, political leaders and parties, and interest groups with substantial financial means. Hence it reduces the spectrum of who can substantially influence policymaking to these who can benefit from AI and big data competitive advantage. Moreover, due to the characteristics of AI, government also introduce a degree of uncertainty and vulnerability when introducing (or allowing the use of) AI in democratic processes. Since the AI-mediation of citizen-government relation remains often blurry if not opaque to the citizen, the "voluntary" aspect of adoption remains debatable. In light of the risks and challenges of AI, there is a dire need to capacity building (i.e., digital skills and literacy) to empower civil society and in particular individual citizens, and to safeguard democratic principles. The EU's approach to trustworthy AI goes in this direction. But this books goes further. It calls for a dedicated approach to AI for the citizen-government relation, and in particular for citizen participation. Each use of AI should apply not only adopt a human-centric approach, but also undergo a risk assessment to ensure the defense of equality, freedom, human rights, and the notion of popular sovereignty.