This deliverable reports on Tasks 2.2 and 2.3 of Work Package 2 (Design AI-enabled Deliberation Methods) of the AI4Deliberation project. It provides a practical guideline for designing deliberative processes that can operate effectively at scale, addressing the central challenge of expanding participation while preserving the quality of collective reasoning. The deliverable translates core deliberative principles — equality, inclusion, plurality, authenticity, and reflection — into implementable design actions for practitioners working with large-scale, digitally supported, or institutionally constrained participatory processes. It combines a conceptual reference framework (Part I) with a modular, constraint-driven application framework (Part II), including deliberative treatments, an AI requirement upscaling matrix, and an operational design tool for upscaling deliberation. A key output is an interactive Shiny-based Deliberative Process Design Tool (DPDT), which allows users to dynamically explore design choices linking deliberative principles, process stages, constraints, and AI-supported options.
Veri et al. (2026) studied this question.