cs.CYMar 17, 2026

Human/AI Collective Intelligence for Deliberative Democracy: A Human-Centred Design Approach

Authors: Anna De Liddo, Lucas Anastasiou, Simon Buckingham Shum

Organizations: Knowledge Media Institute, The Open University, Milton Keynes, UK · Connected Intelligence Centre, University of Technology Sydney, Sydney, Australia

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

This chapter introduces the concept of Collective Intelligence for Deliberative Democracy (CI4DD). We propose that the use of computational tools, specifically artificial intelligence to advance deliberative democracy, is an instantiation of a broader class of human-computer system designed to augment collective intelligence. Further, we argue for a fundamentally human-centred design approach to orchestrate how stakeholders can contribute meaningfully to shaping the artifacts and processes needed to create trustworthy DD processes. We first contextualise the key concepts of CI and the role of AI within it. We then detail our co-design methodology for identifying key challenges, refining user scenarios, and deriving technical implications. Two exemplar cases illustrate how user requirements from civic organisations were implemented with AI support and piloted in authentic contexts.

Figures & tables

Figure 1: Collective Intelligence for Deliberative Democracy Research (CI4DD)
PlatformBCauseDemocraticReflectionKialoPol.isLoomioAdhocracy+Consul
Core Deliberation ModeDialectical (Argument Map)Aggregative (Real-time)Dialectical (Argument Map)Aggregative (Clustering)Consensus (Decision)Modular (Multi-modal)Modular (Multi-modal)
AI roleStructuring & sensemakingSensemaking & facilitationNoneML clusteringNoneNoneNone (planned)
Key featuresTranscript import, arg. network, theme mapsReflection cards, dashboards, AI questionsPro/con tree, sunburst vizStatement voting, group vizThreaded discuss., polls, decisionsIdea submit, budgeting, geo-mapProposals, debates, voting, budgeting
Primary Use CaseCivic engagement, policyLive events, consultationsEducation, debatePolicy-making, large-scaleGovernance, collectivesMunicipal gov., civicMunicipal gov., institutions
Table 1: A Comparative Analysis of Deliberative Technology Platforms
Figure 2: The three-phase co-design methodology: (Community Challenges, Scenarios Co-creation, and Validation) showing key stakeholders (Who), Objectives, and Outputs for each phase in the development of AI-augmented deliberative democracy tools.
Figure 4
Figure 3: “Engaging teenagers” DD user scenario co-created by civic organisations. The scenario follows a structured narrative format: (top panel) begins with Persona description and Context setting, establishing the stakeholder and situation; (middle panel) progresses through Challenge identification and Objective definition, clarifying what needs to be achieved; (bottom panel) details the Actions taken and Outcomes realized, showing how the deliberative technology enables the solution. Coloured annotations on the right link specific narrative elements to system requirements (e.g., UIE.2, DAV.1) detailed in the Appendix, creating traceability between user needs and technical specifications.
Figure 4: Snapshot of BCause Discussion Interface: The focal Question is at the top, and the left panel generates a succinct summary of the most contested and opposed positions . These are listed in the central panel, each with related Arguments organised into Cons (left) and Pros (right). The overview tree on the right supports orientation and navigation.
Figure 5: The BCause policy clusters dashboard, illustrating the main clusters from the ‘Sustainable Food Systems Workshop’. The interface organizes participants into distinct ‘Key Positions’ with each one supported by ‘Supporting Claims’ grounded in a ‘Originating Transcript Context’
Figure 6: Identified argumentative components (claims and premises) are shown as markup in BCause transcript viewer.
Figure 7: BCause interactive clustering analytics page with Voronoi (circle pack) mode selected variable number of clusters
Figure 8: BCause system architecture showing integration of data input sources (face-to-face transcripts, online contributions), AI processing components (Argumentation Mining, clustering, policy recommendation generation), and output interfaces (discussion platform, analytics dashboards, policy reports).
Figure 9: Participant using DemocraticReflection to share their thoughts during a televised leadership election debate. See 3 min introductory video: https://youtu.be/xj0gB07yMuU
Figure 10: Customised Reflection Cards for CEPS Young Thinkers
Figure 11: The DemocraticReflection dashboard for the CEPS event showing two distinct interfaces: (a) a public view with detailed analytical visualizations and (b) a mid-event facilitator view with discussion summaries and engagement metrics.
More in this topic

Deliberative Democracy

Deliberative democracy studies how inclusive, informed public discussion can shape collective decisions and strengthen their legitimacy.

See all 16 papers

Explore similar work

CardsList

  1. AI Should Facilitate Democratic Deliberation at Scale

    Sep 17, 2026José Ramón Enríquez, Jiaxin Pei, Alex PentlandHuman-Centered AIDeliberative Democracy

    Stanford Graduate School of Business, Stanford, CA, USA · Stanford Institute for Human-Centered AI, Stanford, CA, USA · MIT Media Lab, Cambridge, MA, USA

  2. Habermolt: Delegating Deliberation to AI Representatives

    May 23, 2026Joseph Low, Oscar Duys, Claude Formanek +2Deliberative DemocracyAI Agent Governance

    Cooperative AI Research Fellowship · AI Safety South Africa · MIT +1

  3. Accountable Human-AI Deliberation with LLMs: Scaling Collective Intelligence through Symbiotic Scaffolding

    May 26, 2026Wajdi ZaghouaniAI AccountabilityHuman-in-the-Loop AI

    Northwestern University in Qatar