Public participation plays a critical role in AI Impact and needs to be integrated as a necessary component of AI design, deployment and governance. Public participation can address impact in many ways.

• Discover priorities: Inclusive and informed public engagement is needed to understand the impacts that different communities want from AI;
• Co-design and development: Participatory processes are required during the development and evaluation of AI systems to understand how to achieve that impact in inclusive ways; and
• Feedback and accountability: Inclusive participation is required to assess impacts, and to inform the feedback loop between evidence of impacts and policymaker and industry decision making.

Drawing on perspectives from across the Global South (India, Chile, Brazil and Nigeria), and inputs from researchers in the UK, USA, Canada and France, we explored:

• the importance of demonstrating how public participation leads to better AI science;
• the need to create methodologies of participation that “meet people where they are”
• the need for metrics that understand the quality, scope and objective of participatory processes; and
• the value of meaningful inclusion in ‘the last mile’ that engages with bottom-up people-first digital development projects rather than top-down ‘AI-first’ interventions.

Sustainable adoption of AI, including informed choices and the right to reject certain uses, depends on public trust.  But trust cannot exist without participation. Inclusive participation requires understanding how AI, as well as unequal access to AI, affects different groups in different ways. Quantitative public-attitudes research often obscures critical nuance, masking how specific communities are differently impacted by, and feel about, particular AI applications. Meaningful understanding of AI impact therefore requires an ecology of participatory methods, not a single approach.

Public engagement on AI cannot be online-only. Addressing structural inequalities demands sustained, well-resourced outreach to offline communities, people with low digital literacy, speakers of Indigenous languages, and those who rely on intermediaries to participate. Without this, engagement risks reproducing the very exclusions it seeks to address.

Effective public participation helps establish a social contract around technology: shared expectations and obligations between governments, industry, and society to ensure innovation serves the public good. Participatory processes can also