In a webinar hosted by the Brazil Program at the Inter-American Dialogue, policymakers and experts reflected on the outcomes of the AI Summit held in India in February 2026 and on what they signal for the future of artificial intelligence (AI) in the Global South. Moderated by Brazil Program Director Bruna Santos, the conversation featured Beatriz Vasconcellos, deputy secretary for Digital Transformation at Brazil’s Presidency; Luanna Roncaratti, deputy secretary of Digital Government at Brazil’s Ministry of Management and Innovation in Public Service; and Robyn Scott, co-founder and CEO of Apolitical.

The Summit was the first major global AI gathering hosted in the Global South, signaling a shift toward giving emerging economies a stronger voice in shaping the future of AI governance and deployment. While recent debates have focused largely on breakthroughs in frontier models, discussions in New Delhi centered on diffusion and implementation—how countries can effectively deploy AI across government and society.

Speakers emphasized that AI should increasingly be understood as geopolitical and societal infrastructure. Vasconcellos opened the discussion by highlighting that digital technologies—particularly AI—are becoming core infrastructure shaping both geopolitical dynamics and social organization. Echoing a recent remark by President Luiz Inácio Lula da Silva, she noted that algorithms are not merely mathematical code governing the digital world, but components of a broader structure of power. In this context, Vasconcellos framed AI sovereignty for Brazil as a multidimensional concept encompassing four elements: data sovereignty, operational sovereignty, technological sovereignty, and governance sovereignty.

“Digital technologies, and AI as part of them, have become both geopolitical and social infrastructure. I think it’s a concept that those of us working in this field are internalizing little by little. However, this Summit truly highlighted its geopolitical nature, which is also evident in how leaders position themselves.”

Beatriz Vasconcellos

Roncaratti expanded the discussion by introducing the concept of algorithmic institutionalism, emphasizing how algorithms increasingly shape the behavior of institutions, organizations, and individuals. This shift requires governments to rethink how digital systems interact with public administration and policy design.

“We are working extensively [in Brazil] with the concept of national data infrastructure in Brazil as a set of standards, policies, layers, and technological tools, so that we can have more capacity to promote the strategic use of data, especially in the federal government. We know that data needs to be structured and organized, so that AI systems and solutions can activate their potential.”

Luanna Roncaratti

Brazil has been working to strengthen data governance across ministries and improve how government data is organized and shared, recognizing that reliable data is essential for effective AI use. At the same time, she stressed that reducing gaps and expanding AI literacy among both public servants and citizens will be critical to ensuring that the technology benefits society broadly.

Scott highlighted what he described as an AI readiness gap. While political leaders frequently express enthusiasm about AI’s transformative potential, institutions often lack the operational capability to deploy and govern these technologies responsibly. Scott noted that only about 26 percent of public-sector implementers report understanding their government’s ethical frameworks for AI, illustrating the disconnect between policy commitments and implementation capacity.

“This isn’t a technology rollout. This is an institutional transformation.”

Robyn Scott

For Vasconcellos, closing this gap requires governments to develop deeper technical literacy and institutional capacity, allowing them to “understand what is inside the black box” and exercise meaningful oversight over algorithmic systems.

The discussion also addressed the governance challenges posed by AI deployment in the public sector. Roncaratti stressed that responsible adoption requires robust data infrastructure, strong governance frameworks, and thoughtful digital design grounded in ethical principles and secure safeguards.

She highlighted the development of a framework for AI impact assessment in the public sector, designed to evaluate potential risks through classification systems aligned with international standards and governance principles.

Scott added that much of the public anxiety surrounding AI stems from a policy narrative focused heavily on automation rather than augmentation. According to him, discussions about AI’s impact on labor markets often overlook the importance of preserving human agency. Effective governance frameworks, he argued, must ensure that humans remain above the algorithm, with AI serving as a tool to support better decision-making rather than replace human judgment.

Finally, the speakers emphasized the importance of international cooperation in shaping AI governance. Roncaratti highlighted Brazil’s active engagement in global dialogue on digital governance, noting that the country maintains roughly 30 cooperation agreements with other governments and institutions and participates actively in international forums.

Vasconcellos reaffirmed Brazil’s longstanding commitment to multilateralism, arguing that building common frameworks and shared norms will be essential for managing the global implications of AI. Collaborative approaches, she noted, allow countries to exchange knowledge, align governance principles, and develop collective strategies to address emerging technological challenges.

Scott concluded by noting that there remains a significant—and largely untapped—opportunity for Global South collaboration around institutional readiness, governance frameworks, and capacity building. Strengthening these partnerships could help countries collectively navigate the technological and policy transformations driven by AI.

WATCH THE EVENT HERE: