This recipe is about designing an entire democratic process—not just the AI tool within it. When AI is introduced into a deliberative setting, the surrounding process needs to change too: not just to make the AI work, but to make sure the democracy works.

It offers practical steps to help you embed AI into a participatory or deliberative process in a way that maintains transparency, critical distance, and participant agency. That means thinking carefully about how and when participants engage with the AI, how they are prepared to do so, and how the rest of the process enables them to reflect, challenge, and reframe the technology’s outputs.

This approach was tested in the Digital Democracy Lab, where citizen participants were guided through profiling, data source selection, prompt generation, and critical review. Rather than hiding complexity, the process exposed how the AI worked—and gave participants the tools to question it.

This recipe is part of the Digital Democracy Lab Handbook — a practical resource for facilitators, technologists, and public sector actors exploring how AI can support meaningful democratic exchange.