Explainable AI for Smart Decision-Making in Swiss Referendums (smartinfo)
The idea behind the online voting aid “smartvote” is to be applied to referendums. An interdisciplinary research team is investigating how AI can help improve information and participation among the Swiss population.
Factsheet
- Schools involved Business School
- Institute(s) Institute for Public Sector Transformation
- Research unit(s) Public Sector Innovation
- Strategic thematic field Thematic field "Humane Digital Transformation"
- Funding organisation Others
- Duration (planned) 01.01.2026 - 31.12.2028
- Head of project Dr. Daniel Schwarz Badertscher
-
Project staff
Dr. Daniel Schwarz Badertscher
Jan Fivaz -
Partner
Hochschule Luzern
Année Politique Suisse, IPW
Dachverband Schweizer Jugendparlamente / Easyvote
Politools - Keywords Voting advice application, referendums, information dissemination, participation, explainable generative AI, deliberation
Situation
Switzerland is the country of direct democracy. The average turnout in popular votes, however, is less than 50%, and a significant number of citizens are unable to provide a content-related justification for their vote choice. An interdisciplinary team of researchers from the BFH, the Lucerne University of Applied Sciences and Arts, the University of Bern, the "easyvote" project of the Organization of Swiss Youth Parliaments, and the "smartvote" project of Politools, are looking for ways to enable citizens to make informed decisions without having to study the often complex proposals in detail. A modular platform is being developed that combines conventional information, AI-supported elements, and real interaction between citizens. Particularly, the AI module is investigating how XAI (explainable AI) and GenAI (generative AI) can help improve informed and inclusive democratic participation in referendums. The aim is to create a platform that helps voters to form an informed opinion and at the same time promotes exchange among citizens. To this end, generative AI is used that meets comprehensive criteria of transparency, traceability, and compliance with data protection regulations. The long-term goal is to create a new, trustworthy source for political opinion formation.
Course of action
Conception, development, and evaluation are carried out in close cooperation between computer science and social science researchers and the participating NGOs. External experts from various fields are also involved, who assess the project at the beginning and the results achieved at the end of the project. In addition, test runs and scientifically based surveys are conducted among the target groups.
