Participating with AI: how generative visualisations are changing participatory planning

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

This study examines how generative visual AI – tools that transform photographs of real places through text-prompted image generation – changes participatory urban planning in practice. It analyses five early cases (2023–2024) in which the pioneering UrbanistAI tool was integrated into official participation processes in Finland, Kosovo, and the United Arab Emirates, drawing on eight interviews with planners and consultants and on qualitative content analysis. Under participatory AI, the article distinguishes participation with AI – the use of AI tools within participatory activity – from participation in AI, public involvement in the governance of algorithmic systems. The findings show that immediate, iterable visualisations improved communication, clarified abstract ideas, and strengthened iterative co-design and joint evaluation within workshops. The tool did not replace professional expertise or planning procedure but functioned as a low-threshold instrument for expression and deliberation. Its influence on built-environment outcomes depended on whether participatory inputs were connected to an institutional feedback loop with the scope to act on them. Theoretically, the article proposes the designed distribution of agency as a way of specifying how a given tool arranges the relationship between human and machine – admitting machine co-agency in an imaginative phase while re-concentrating accountable judgement in a human-controlled design phase.