Informal Stakeholder Consultation on the Global Dialogue on Artificial Intelligence Governance

Tim spoke as part of the Informal Stakeholder Consultation on the Global Dialogue on Artificial Intelligence Governance.

Tim’s speaking notes are included below:

Thank you to Susan Oman from the University of Sheffield for sharing this speaking slot. I speak as a members of the Programme Committee for PAIRS - the Participatory AI Research & Practice Symposium - a community of researchers and practitioners focussed on bringing voices from diverse publics across the world into the governance of AI.

We thank co-chairs for the opportunity to input into the design of the Global Dialogue.

We echo the calls from Antoine, Marine and others earlier on the critical need for the Global Dialogue to prioritise the direct representation and engagement of the voices of everyday citizens and of affected communities - and the need for the events in July to facilitate two-way dialogue with citizens: establishing mechanisms through which grassroots deliberations and public engagement to feed into and be made visible at the Dialogue - and through which the expert discussions at the Dialogue can inform not just global discussions, but also distributed, national and subnational citizen action on AI governance.

Building on the work of the Global Citizens Assembly that docks into the Climate Conference of Parties, on discussions at PAIRS alongside the India AI Impact Summit, and learning from work such as the We The Internet project Antoine mentioned earlier, we are working on a process over the next few months to bring together partners to articulate design options for a Citizens Track that could deliver an effective two-way interface between the Global Dialogue and the diversity of models of deliberative engagement at local, national and transnational levels.

At this early stage of that design process we anticipate there are a number of ways the design of the Global Dialogue will be able to support this:

• In terms of content: The dialogue should take stock of existing public engagement and public attitudes work on AI. This should provide space for evidence from robust public attitudes research and citizen engagement - including engagement with children and young people, and older people - to be input alongside the scientific evidence to ensure priorities are shaped through robust evidence on public concerns. It will also help highlight gaps in public voice - surfacing areas where more investment and action is needed to hear from citizens across different geographies and communities.

• In terms of enabling inclusive participation we encourage thought to providing physical space for public engagement and th visible presence of citizens affected by AI. Alongside the first AI Safety Summit in the UK in 2023, my organisation Connected by Data, brought together a small peoples panel of sortition-selected citizens who followed debates in real-time, and had a visible presence in fringe events. Having affected citizens in the room can help shape debate, and creating supportive space for direct citizen representation could form a valuable part of supporting a Citizens Track.

• And in terms of outputs, we believe it would be useful to explore how outputs can include Framing questions for public engagement. The dialogue can have a role in articulating concrete issues for grassroots public engagement in the period between dialogues. Questions and expert evidence coming out of the Dialogue, and clear invitations for public input, can support deliberative practitioners to focus their efforts.

We look forward to the opportunity to input written work as our collaborative design process on this work produces outputs over the coming months.