BRÉSIL PERSPECTIVE

Brazil approaches AI regulation from its position as an emerging giant torn between Global South leadership ambitions and the reality of its technological dependence. Folha de Sao Paulo and O Globo cover the American legislative framework by systematically comparing it to the Projeto de Lei 2.338/2023 — the Brazilian 'Marco Legal' approved by the Senate in December 2024, transferred to the Chamber of Deputies in March 2025, but stalled since. The Brazilian AI Plan (PBIA) 2024-2028, titled 'AI for the Good of All,' displays considerable ambitions: 23 billion reais for national computing infrastructure, sovereign cloud capabilities, and Portuguese-language foundation models.

Brazilian framing is deeply shaped by BRICS logic and the 'Global South.' The Estadao notes Brazil rejects both the American deregulation model — perceived as submission to Silicon Valley interests — and Chinese control — perceived as authoritarian. The Marco Legal adopts a risk architecture similar to the EU AI Act but with particular emphasis on civil liability and a National System for AI Regulation. However, researcher Gabrielle Graca of Fundacao Getulio Vargas tempers optimism: the bill has little chance of passing in 2026, an election year in Brazil where technological debates always yield to social emergencies.

The major omission in Brazilian coverage is the abyssal gap between stated ambitions and actual capabilities. The PBIA's 23 billion reais represents a fraction of US and Chinese investments, and Portuguese-language model development remains embryonic. Amazonian sovereignty — Brazil's media reflex against international pressure — transposes to digital: 'data sovereignty' is brandished as a flag without Brazil possessing the infrastructure to exercise it.