Innovation or Exclusion? Evaluating the Role of Artificial Intelligence in Reducing Social Inequality in Brazil

AI is becoming central to Brazil’s innovation and modernization strategies, yet its social impact is deeply stratified. National policies, such as the Estratégia Brasileira de Inteligência Artificial, aim for inclusive outcomes; however, access remains limited, particularly in rural areas, where over 40% of the population lacks consistent internet. While urban centers experiment with AI in public services, marginalized communities remain disconnected from both the infrastructure and benefits.

The current research literature contains two distinct, opposing views. The Inter-American Development Bank, in collaboration with UNESCO and other international organizations, demonstrates how AI serves as a democratizing force across Latin America by utilizing adaptive learning platforms, diagnostic algorithms, and innovative job-matching tools to enhance access to education, improve healthcare, and enhance labor market efficiency. These technological solutions are often presented as effective remedies that can address deep-seated social inequalities.

Research conducted by IPEA and scholar R. Leal, along with other critical studies, suggests that AI systems will exacerbate social inequalities when structural protections are absent. Training datasets containing biases tend to reproduce social order by giving preference to users who are urban, wealthy, and light-skinned. AI benefits primarily accumulate in economic centers like São Paulo because unequal distribution of digital infrastructure prevents technological access for vast rural and peripheral regions. The Brazilian private sector dominates AI development because profit-driven goals often outweigh concerns about equity, leaving public institutions unprepared to use AI responsibly.

Brazil’s Estratégia Brasileira de Inteligência Artificial presents forward-thinking language about inclusive and ethical AI development. However, implementation remains fragmented, and the gap between policy and practice is particularly evident in public services. Pilot AI projects in education or health often fail to scale beyond urban testing zones or reach under-resourced regions. While global literature focuses heavily on AI ethics, algorithmic bias, or labor automation in wealthier countries, few studies address how AI functions in societies where inequality is not a side effect but a foundational structure.

This study shifts the lens from “Can AI work?” to “Who benefits from AI, and who continues to be left out?” It aims to uncover the invisible conditions of political inertia, uneven infrastructure, and racialized access that determine whether AI can become a transformative tool or just another layer in Brazil’s digital hierarchy.