Political opinion formation is at the heart a functioning democratic society. However, it has been subject to major changes in recent years: In addition to the traditional media, various actors are now determining public discourse, which is often influenced by one-sided filter bubbles and echo chambers. For citizens, this results in uncertainty, confusion and, not least, polarization. However, whether and who they vote for depends largely on the information they consume in their everyday lives.
The information landscape is split between a number different content providers - social media, influencers, messaging services and many more. Information is often presented without reference to sources or context. Recipients cannot verify this information in the same way that they are able to verify information provided by traditional journalism. The floodgates are open to misrepresentation, manipulation, distortion or artificially created content. As a result, democratic processes, such as political opinion formation and free elections, are disrupted, existing views are reinforced and constructive discourse is undermined.
This is where the new research alliance funded by the Daimler and Benz Foundation comes in. In the project, an innovative software tool (KonCheck) is to be developed and tested. Taking the form of a user-friendly app, it will be able to contextualize politically relevant information and to check it for authenticity and trustworthiness. To this end, artificial intelligence language models will be employed, enabling users to ask questions about texts, view sources, retrieve articles in simple language or have context provided. KonCheck is intended to encourage participation by providing easy opportunities for interaction.
The AI tool will be designed to be intuitive and easy to understand, as it is primarily aimed at vulnerable user groups. These include people with cognitive impairments, senior citizens as well as young first-time voters. As so-called digital immigrants, elderly people may sometimes share fake news due to a lack of media literacy. Due to their early consumption of social media, young voters, on the other hand, are often socialized in polarizing information environments and may be less willing to compromise.
The interdisciplinaly reserach alliance on "AI-based methods to support opinion formation and participation" is led and coordinated by Professor Dr. Jens Gerken, head of the research group for "Inclusive Human-Robot Collaboration" at TU Dortmund University. The development of the new AI tool is to be informed by perspectives from social psychology and communication science. The project aims at supporting people in their personal decision-making processes and strengthen their resilience to manipulation. Political participation is an expression of social justice and strengthens democratic mechanisms.
Project partners
- TU Dortmund University
- Professor Dr. Jens Gerken (coordination and project lead), Inclusive Human-Robot Collaboration
- Professor Dr. Jürgen Howaldt, Social Research Centre
- Professor Dr. Johannes Weyer, Social Research Centre
- Professor Dr. Matthias R. Hastall, Communication for Health, Inclusion and Participation
- CISPA Helmholtz Center for Information Security
- Professor Dr. Thorsten Holz
- Westfälische Hochschule
- Professor Dr. Norbert Pohlmann, Institut für Internet-Sicherheit
- Rhein-Waal University of Applied Scienes
- Professor Dr. Patrick-Benjamin Bök, Cloud Resilience Lab
- Professor Dr. Matteo Große-Kampmann, Cloud Resilience Lab
About the Ladenburg Research Networks
The Ladenburg Research Networks are a major funding program of the Daimler and Benz Foundation. This format offers scientists the opportunity to work on topics within an interdisciplinary research network over an extended period. For this purpose, the Foundation publishes calls for applications at irregular intervals. For more information, please visit www.daimler-benz-stiftung.de
