We are excited to announce the release of our new working paper, now available on arXiv. In this new working paper, Adrian Rauchfleisch and I continue our exploration of people’s opinions on Artificial Intelligence (AI) in democracy and politics. After an earlier study on the public’s views on AI in campaigning, we now turn to public opinion on AI in deliberation:
Digital deliberation significantly expands opportunities for democratic engagement, yet managing deliberative processes at scale remains challenging. While AI promises to enhance efficiency in digital deliberation, our findings reveal a noteworthy “AI penalty”: public skepticism towards AI in deliberative contexts.
Through a preregistered survey experiment with a representative sample in Germany (n=1850), we discovered that participants exhibit less willingness to engage in AI-facilitated deliberations compared to human-led formats and rate their quality significantly lower. Crucially, attitudes towards AI—including perceived benefits, risks, and the extent of anthropomorphization—significantly moderate these effects.
Our study highlights the emergence of a new deliberative divide driven by individual attitudes towards AI. This divide presents a novel challenge to democratic participation, distinct from traditional divides shaped by demographics or educational background. As democratic practices increasingly move online and leverage AI, understanding and addressing public perceptions and hesitancy toward AI will be critical.
Read our full working paper here: https://arxiv.org/abs/2503.07690
Andreas Jungherr, and Adrian Rauchfleisch. 2025. Artificial Intelligence in Deliberation: The AI Penalty and the Emergence of a New Deliberative Divide. arxiv. Working Paper. doi: 10.48550/arXiv.2503.07690