<aside> ℹ️ This presentation features at PAIRS 2026 in New Delhi on 18th February 2026 at 11:00 IST
🎙️ Thread on PAIRS Discussion Server (Discord) (register first)
</aside>
The rapid evolution of artificial intelligence (AI) systems—coupled with deep uncertainty about their societal impacts and competing visions of what constitutes inclusive and meaningful governance—presents an urgent global challenge. Developing effective AI governance frameworks requires more than technical expertise; it demands navigating complex issues at the intersection of technology, society, and public accountability. These challenges unfold amid rising political polarization, declining public trust, and persistent socioeconomic inequalities.
In many emerging economies across the Global Souths, AI governance must coexist with pressing imperatives to deploy AI for social development. This tension—between adopting global governance standards and safeguarding local priorities—raises critical questions about how to design inclusive, context-sensitive approaches that avoid reproducing technological or epistemic colonialism.
Deliberative and participatory methodologies offer valuable pathways to address these tensions by engaging diverse publics across power and informational asymmetries. This proposal presents a comparative case study of two participatory processes in the development of AI governance frameworks in Latin America: the elaboration of Chile’s National AI Strategy (2019-2021) and the adoption of Brazil’s AI Code of Conduct in the Legal System (scheduled for early 2026).
Case 1: Deliberative Polling for the use of AI in Brazil’s Legal System (the Brazilian AI Act)
Organized by the Stanford Deliberative Democracy Lab, in collaboration with Brazil’s Institute for Development and Research (IDP) and the Order of Attorneys of Brazil (OAB), this deliberation will take place in early 2026 right after the congressional vote on Brazil’s landmark AI Bill. The methodology—previously deployed by the Lab in over 100 projects worldwide—uses a deliberative polling format structured as follows: (1) expert body that map proposals and come up with proposals for the deliberation (in this case, paying special attention to what safeguards are needed for AI to be used in the legal system in a way that goes in accordance with Brazilian law and the AI bill) (2) selection of the deliberating body; (3) opening plenary; (4) pre-survey; (5) small-group deliberation; (6) expert exchanges; and (7) post-survey.
The deliberative body will consist of a random sample of 150-200 lawyers who are members of the Order of Attorneys of Brazil, who will deliberate in person on specific elements of the National Council of Justice’s proposals for the implementation of AI in the courts, as well as proposals put forth by nonprofits, and other stakeholders, including provisions related to data privacy, high risk use, and human rights. Through comparative analysis of the pre- and post-surveys, the process will capture participants’ evolving perceptions, beliefs, and regulatory priorities. The most salient findings and recommendations will be incorporated into the final version of the national code of usage of AI in the legal system in Brazil.
Case 2: Participatory Crowdsourcing for Chile’s National AI Strategy
In 2021, amid nationwide social mobilization, a constitutional re-drafting process, and the COVID-19 pandemic, Chile launched its National AI Strategy. Responding to calls for more transparent and inclusive governance, the Chilean government implemented a series of participatory consultations in 2020—both in-person and virtual—across the country. People gathered in self-organized roundtables and regional workshops to deliberate about the socioeconomic priorities shaping Chile’s AI agenda.
By comparing these two participatory AI governance models, this study examines how informal, adaptive approaches to participation (as in Chile’s AI Strategy) contrast with more structured deliberative methodologies (as in Brazil’s AI Code of Conduct). The analysis highlights the trade-offs between flexibility and procedural rigor, and explores whether more formalized participatory models can enhance inclusiveness, legitimacy, and policy coherence without constraining local input or responsiveness.
Foregrounding empirical evidence from Latin America, this study contributes a vital Global Souths perspective to the emerging field of participatory AI governance. It shows how context-specific political cultures, institutional capacities, and civic traditions shape the meaning and practice of participation in AI policymaking. These insights are not only regionally significant but also offer valuable lessons for other emerging economies—such as India—where balancing technological leapfrogging with democratic accountability remains a central challenge. In doing so, this work advances a more plural and comparative understanding of AI governance, centered on community power, participatory practice, and democratic decision-making.