The Impact of Artificial Intelligence on Democracy: A Challenge for Researchers
According to the OECD, the World Economic Forum, and several other international organizations, generative artificial intelligence is one of the most significant forces shaping societies in the 2020s. The AI industry is not only producing new software for consumers —it is also transforming labor markets, the economy, and future skill requirements. Algorithms are now capable of influencing political systems, societal stability, and democratic development.
Researchers are urgently expected to provide thorough analysis of the impacts of generative AI. In support of decision-making and foresight, we need answers to questions such as how algorithms reshape participation, willingness to cooperate, and other mechanisms that uphold democracy.
From a research perspective, studying the democratic effects of evolving technologies is a challenging task. These effects are complex and influenced by multiple factors. It is not enough to examine how democratic institutions or their sustaining mechanisms and practices are changing. Researchers must also identify to what extent these changes can be attributed specifically to generative AI.
The challenge is further compounded by the current lack of reliable methods or metrics for measuring the impacts of generative AI. Developing such methods is difficult when there is so little concrete knowledge about its actual societal effects.
As researchers from Carnegie Mellon University recently noted, the academic literature currently focuses primarily on projected rather than observed impacts of generative AI (Lai et al., 2025). Projected effects — such as assumptions about potential risks — may offer valuable perspectives. However, they also introduce significant uncertainty.
At AIdemoc, we are actively seeking methods for researching the impact of generative AI on democracy. Our project combines computational, empirical, and foresight-based approaches across disciplinary boundaries. We examine the democratic implications of AI both in the short term (by 2030) and long term (by 2050). We also explore how AI may influence democracy through both intentional (e.g. communication capture) and unintentional (e.g. language model hallucinations) contamination of the information environment, and we search for ways to prevent such contamination.
Additionally, we study how digital democracy can be strengthened and how generative AI — such as language models and speech synthesis—can enhance the inclusion of underrepresented groups, especially older adults and non-native speakers.
AIdemoc is part of the Strategic Research Council’s DEMOC program, running from 2024 to 2030. The project consortium includes the University of Helsinki, VTT Technical Research Centre of Finland, and the University of Jyväskylä. Project interaction is coordinated by Sherpa. Partners include government agencies, ministries, education, media and communications organizations, and civil society groups.
