Institutionalised Distrust and Human Oversight of Artificial Intelligence: Towards a Democratic Design of AI Governance under the European Union AI Act

This course examines human oversight as a key tool in AI governance, highlighting empirical evidence that humans often fail in oversight roles due to limited competence or harmful incentives. It analyzes the EU Artificial Intelligence Act, proposes a taxonomy of oversight roles, and develops six normative principles — drawn from democratic theory — to institutionalize distrust and strengthen the effectiveness and trustworthiness of AI oversight.