How can artificial intelligence contribute to more constructive public discourse and a stronger democratic culture? This question drives the research project Human–Machine Interaction for Democracy (HMI4D), funded by the Mercator Switzerland Foundation and the Hasler Foundation.
The project brings together an interdisciplinary team from the Institute for Business Ethics (IWE-HSG), the Institute of Computer Science (ICS-HSG), Pro Futuris Think + Do Tank, the Center for Democracy Aarau (ZDA), and Beobachter. By combining expertise in ethics, political science, and computer science with hands-on experience in dialogue formats, the initiative seeks to bridge the gap between theoretical research and practical democratic innovation.
At its core, HMI4D develops an AI-based conversational interface called FRANCA (Free Rational Application of Normative Communicative Action). Drawing on deliberative democratic theory and discourse ethics, FRANCA focuses on how people reason together rather than on what they say. It evaluates key dimensions of dialogue quality, such as justification, reciprocity, respect, and engagement with counterarguments, and provides participants with structured feedback to support more reflective and balanced exchanges.
A central innovation of the project is its focus on everyday interactions. Whereas most existing initiatives target large-scale deliberative forums or “mini-publics,” HMI4D addresses one-on-one and small-group discussions, where democratic reasoning most commonly occurs. By embedding AI directly into these settings, the project aims to foster a more continuous and socially grounded practice of reason-giving and mutual understanding.
FRANCA is designed as a facilitative tool rather than an authoritative system. It does not evaluate participants or prescribe outcomes. Instead, it encourages reflection by offering feedback on the dynamics of the conversation, supporting meta-dialogue and learning processes. In this way, it helps individuals refine how they argue, listen, and engage with opposing perspectives, strengthening the normative foundations of democratic discourse.
By integrating philosophical insights, empirical research, and technological development, HMI4D advances a new interdisciplinary approach to studying and shaping democratic communication. Its ambition is not only to analyse how AI influences reasoning, but to actively design systems that improve the quality of democratic life.