Towards a Citizens Track on AI Governance

How can the voices of affected communities be placed more central to AI governance?

Building on discussions at the Participatory AI Research & Practice Symposium (PAIRS) alongside the India AI Impact Summit, and with a view towards the UN Global Dialogue on AI Governance, and next AI Summit, both due to be held in Geneva in 2026 and 2027 respectively, we want to explore practical approaches to centre citizen voice in global AI governance fora.

In partnership with the Democracy Section, Peace and Human Rights Division of the Federal Department of Foreign Affairs, Switzerland, ISWE Foundation and Connected by Data hosted a two hour online workshop as the first of a series of convenings in the run up the UN Global Dialogue to explore:

• What we collectively know about plans for the Global Dialogue and AI Summit Series: building on experience of past Summits, and emerging plans for 2026/27.

• Existing participatory AI practice: drawing on work shared at PAIRS and other fora, and past and prospective global deliberations on data and AI, to explore inspirations for an AI Citizens Track

• Learning from the Citizens Track on Climate launched at COP30 last year.

• Practical workstreams to develop a citizens track presence in Geneva in July 2026: position paper, evidence compendium & expo presence.

Session Report

Thank you to all 50+ people who joined us for the kick-off meeting on Monday. It was fantastic to see the diversity and energy of the group coming together.

Next steps:

To keep in touch, and get involved in the next steps of this work, please fill in this quick form which also asks about interest in the three action areas we discussed:

• Position paper - working towards the next step of a workshop session on 23rd April (Hybrid meeting: Geneva & online) to develop the draft that will address the Why? What? and How? of the Citizens Track as an interface between the UN Global Dialogue on AI Governance (and other AI governance processes), and existing distributed and transnational citizen deliberation.

• Case book - gathering examples of existing participatory practice - and supporting the case for a standing mechanism that can amplify, complement and connect existing and upcoming public deliberations on AI

• Expo presence - exploring the idea of making public participation on AI visible within the AI for Good summit taking place alongside the UN Global Dialogue in July.

You can find the form here.

Session notes and summary

The shared notes document from the call is available here. We heard from:

• Iñaki Goñi on learning from People in the Lead, and the development of a Citizens Track in the climate process, running parallel to the UNFCCC in an annual cycle, and with impact measured through metrics on: institutional actions; citizen action; solidarity; learning at scale and inclusion.

• Chinasa T Okolo on the UN Global Dialogue on AI Governance currently in development with an emphasis on international cooperation, sharing best practice and inclusive process. Chinasa pointed to the current consultation call for written input, and explored links between the dialogue and the Independent International Scientific Panel on AI.

• Jovan Kurbalija on the importance of centering knowledge, closing the feedback loop, and exploring policy totems that people can gather around, using the example of dialogue tablecloths. Jovan also emphasised the importance of thinking about subsidiarity.

The call included six inspiring examples of participatory practice at different levels, including:

• Arpita Kanjilal on reaching grassroots communities
• Pierre Noro on consultations ahead of the French AI Action Summit.
• Antoine Verge on the Coalition for Inclusive AI, and the We the Internet project
• Faisal Lalani on the Collective Intelligence Project’s work on Collective Input; Collective Evaluation and Collective Intelligence.
• Canning Malkin on the global assembly on climate and food security that has just completed its transnational dialogue sessions.
• Jose Antonio Guridi Bustos on public engagement with AI policy making in Chile

Tim Davies presented three areas for action and invited participants to join breakout discussions to explore how the ideas put forward could be made achievable, impactful and inclusive.

• A position paper, making the case & providing design for a citizens Track. Aiming for:
  • Paper by July 2026
  • Early inputs to UN Global Dialogue consultation by end April
• A case book: collating existing evidence and proof points
  • Building on PAIRS & PVAI Evidence Review
  • Capturing public messages, mapping field, sharing methods
• Geneva presence: Expo or other events
  • Exploring opportunities to make public participation more visible at Dialogue/AI for Good
  • Exploring opportunities for a Citizens Hub

The shared notes document contains brief reflections from these breakouts, and the sign-up form provides a way to indicate interest in working on these ideas further.

Feedback themes

Discussions on zoom, and in the meeting chat, surfaced a range of themes for the Citizens Track design work to consider, including:

• Language of citizenship: Participants debated the use of the word “citizen,” noting it carries “blood and soil” baggage in some contexts and can be exclusionary toward immigrants and refugees. Alternatives like “community science” or reclaiming the term through “global citizenship” were suggested.
• Global South & Non-Western Views: There was a strong call to involve communities in the Global South, particularly where data centers are being built, to ensure their worldviews, ecosystems, and non-Western knowledge are not sidelined.
• Incentives and Accessibility: Participants noted that trust is situational and that participation must be incentivized. Without careful design, “public” engagement may only reach those with existing interest, time, and financial bandwidth.
• Terminology as a Barrier: Engagement often uses technical terminology that alienates communities; participants urged a focus on listening to how technology affects people in their own language.
• Technical Framing: AI is often narrowly perceived as the domain of IT specialists, which limits engagement from professionals in sectors like labor, education, and ethics.
• Multistakeholderism vs. Participation: A primary concern was that “multistakeholderism” often empowers only those with money and specialized knowledge rather than the broader public.
• The Role of Scale: Discussion occurred regarding whether “scaling” participation follows a “Big Tech” logic. Some argued scaling should be “fractal”—reaching more people while remaining deep and meaningful—as a means to shift social power dynamics.
• Institutional Skepticism: Participants flagged skepticism toward the UN, citing past instances where critical “AI for Good” keynote speeches were reportedly censored.