Produced in collaboration with the Organisation for Economic Co-operation and Development (OECD), Artificial Intelligence and the Future of Citizen Participation examines 50 case studies from 22 countries.
The research shows how governments and civil society organisations are already employing AI to analyse public feedback, moderate discussions, answer questions and help more people take part in democratic processes.
One of AI’s most valuable abilities is making sense of large quantities of information. Public consultations can attract thousands of comments, suggestions and personal accounts, creating a considerable administrative task. AI systems can organise these contributions, recognise recurring themes and produce summaries for human officials to examine.
This could allow public bodies to understand people’s opinions more quickly without overlooking contributions simply because too much information has been submitted.
The study identifies nine principal uses for AI in citizen participation: sense-making, information development, translation, transcription, virtual assistance, facilitation, moderation, simulation and participation architecture.
Sense-making, automated translation and virtual assistance are currently among the most widespread. Translation and transcription can remove some of the barriers preventing people from participating, while chatbots can help residents locate information and discover how to contribute to consultations.
There are already encouraging examples. In Cambridge, an AI tool was used in 2024 to help analyse qualitative responses to a public consultation. It reportedly reduced the time required by approximately 50 per cent.
In Finland, the innovation agency Sitra used the AI-powered Polis platform during its “What do you think, Finland?” consultation. More than 18,000 people took part.
For governments, and for businesses supplying technology to the public sector—the potential opportunities are considerable. AI could enable participation exercises to reach larger audiences without producing an unmanageable administrative burden.
However, the report also highlights some serious risks.
Biased training data could distort the interpretation of public opinion, while systems that cannot explain how they reached their conclusions may undermine confidence in the process. People without reliable internet access or adequate digital skills could also find themselves excluded.
Public bodies may lack the specialist knowledge needed to evaluate AI systems properly. There is also the danger of becoming dependent on proprietary platforms controlled by individual technology suppliers.
The Policy Brief recommends action in three broad areas: guardrails, enablers and public engagement. Governments should establish clear rules for the use of AI, develop internal expertise and involve citizens in the design and supervision of participation systems.
The central message is that AI should support democratic judgement, not replace it. Technology may be able to sort thousands of comments or translate contributions almost instantly, but decisions must remain accountable to people.
Used carefully, AI could help governments listen to more voices and respond more effectively. Used without sufficient transparency or oversight, it could damage the very trust that citizen participation is intended to build.
The Policy Brief, Artificial Intelligence and the Future of Citizen Participation, can be downloaded from the Bertelsmann Stiftung website, here:- https://www.bertelsmann-stiftung.de/en/publications/publication/did/artificial-intelligence-and-the-future-of-citizen-participation
No comments:
Post a Comment