Abstract

Artificial intelligence (AI) presents both opportunities and risks for citizen data. AI-enabled tools, when used appropriately and safely, can facilitate the collection, processing, classification, translation, analysis and dissemination of data, potentially reducing costs, expanding analytical capacities and making community-generated evidence more accessible and useful. However, these benefits come with significant risks, including algorithmic bias, misrepresentation of lived experiences, privacy violations, lack of transparency and the reproduction of existing inequalities. These concerns are particularly important when citizen data reflect the experiences of marginalized groups or address sensitive social realities.
The relationship between AI and citizen data also works in the opposite direction. Citizen data can help identify biases and blind spots in AI systems, challenge dominant narratives and bring community knowledge and lived experiences into the assessment of AI-generated outputs and decisions. By giving agency to citizens and communities, documenting disparities, monitoring impacts and providing evidence grounded in people's realities, citizen data can contribute to greater transparency, accountability and fairness in AI development and use. Realizing this potential, however, requires attention to data quality, representativeness, community agency and the risks associated with how data are collected, interpreted and used.
The presentation will explore these opportunities and challenges through the 13 principles of the Copenhagen Framework on Citizen Data. It will consider how principles including independence, relevance, participation and inclusion, informed consent, professional standards, data security, and self-definition and self-identification can guide the responsible use of AI in citizen data initiatives, while helping ensure that citizen- and community-generated evidence can contribute to more ethical and accountable AI systems. It will argue that responsible innovation requires more than technically sound data and algorithms: it also requires meaningful participation, transparency, protection from harm and accountability to the people whose lives the data represent. Ultimately, the presentation will examine how the principles of the Copenhagen Framework can help ensure that AI strengthens rather than undermines the integrity, agency and public value of citizen data, and that citizen data serve not only as an input to AI systems but also as a means of holding them accountable.