Brazil Accelerates Adoption of Artificial Intelligence
Institute for Statistical Studies and Economics of Knowledge has analyzed the key statements of Brazil’s plan for artificial intelligence (AI) development approved by government, which envisages intensifying of research and training highly qualified personnel as well as AI technology implementation expanding in main industries.
Recently Brazil demonstrates significant progress in AI development: according to AI Index by Stanford University, it rose 19 positions – from 35th to 16th place in 2023–2024. Its highest positions are in sub-indices "Talents" (12th), "R&D" (15th) and "Economic integration" (18th). At the same time in terms of "Infrastructure" and "Management" Brazil occupies only 26th and 32nd positions accordingly. A shortage of computing power hinders national scaling of large language models and performing of resource-intensive calculations, and lack of regulation in AI area hampers introducing of AI-based solutions to critical industries, such as healthcare and finance.
To improve balance of regulation, public funding priorities and personnel training in AI area, Brazilian Parliament approved Plano Brasileiro de Inteligência Artificial (BAIP) in November 2024.
Reference:
Brazilian Artificial Intelligence Plan (BAIP) is a strategic initiative of the National Council for Scientific and Technological Development (CNPq) of the Federal Government, under coordination of the Ministry of Science, Technology and Innovation (MSTI), with support of the Center for Management and Strategic Studies (CGEE). The Council provides advisory support to MSTI under quarterly meetings and since 2022 it has been granted the highest advisory body status to the President on science, technology and innovation (STI) policy. The Council's main functions are to develop, monitor and evaluate national STI strategies, including aligning sectoral programmes with the New Industry of Brazil Programme objectives and the National Strategy for Science, Technology and Innovation 2022–2032. The Council's activities are carried out through thematic working groups focused on digital transformation, bioeconomy, “green” industry and strategic technologies. The Council comprises 28 members, including 12 representatives from federal ministries (STI; education; health care; defense, etc.), 8 from academic community, 5 from private sector (including industrial and ICT associations), as well 3 from regional and municipal STI agencies.
Source: ISSEK HSE according to CNPq.
The document targets to stimulate AI technologies implementation in all economy sectors and envisages allocation of R$ 23.1 billion (RUB 330.1 billion) for the period 2024–2028. Funds will be distributed across 14 programmes within five areas. The largest amount of funding (R$ 13.8 billion / RUB 197.3 billion) is allocated to support industry. Next by specific weight are investments of R$ 6.2 billion (RUB 88.6 billion) to science infrastructure and R&D; R$ 1.8 billion (RUB 25.7 billion) to data networks creation, and R$ 1.2 billion (RUB 17.2 billion) to talents development.
Reference:
The financial sources for BAIP performance are the following: 4.6% - private sector; 1.9% - regional budgets; 93.5% - federal budget, including 12.6% - grants of CNPq and Federal Agency for Support and Evaluation of Graduate Education (CAPES), 56.3% - grants and concessional loans from Brazilian Development Bank (BNDES) in a partnership with Funding Authority for Studies and Projects (FINEP) and National Fund for Scientific and Technological Development (FNDCT), 0.4% - BNDES repayable (concessional) loans.
Source: ISSEK HSE according to CNPq.
Wide range of target indicators are envisaged under BAIP, including: doubling volume of business investment to AI by 2028 and 25% increasing share of enterprises implementing AI technologies to production processes; ensuring country’s position in top five global ranking by computing power; introducing of Brazilian AI governance and regulation model by at least five developing countries (tab. 1).
Table 1. Activity and Target indicators of Brazilian Artificial Intelligence Plan
Source: ISSEK HSE according to CNPq
Research infrastructure and grants
More than half of envisaged funds R$ 1.8 billion (RUB 25.7 billion) for domestic AI equipment development will be allocated in 2026 to purchase 5000 graphic processors (GPU). This computing power is a basis for supercomputer creation, which components are designed jointly by Brazilian research and industry organisations with allocation of R$ 765 million (RUB 10.9 billion) and target launch by 2028. Universities and research organisations will have a priority access to supercomputer. R$ 250 million (RUB 3.6 billion) are allocated to related digital infrastructure development. Despite the relatively small scale of the computing installation, its integration into international computing network is planning in the framework of intergovernmental S&T agreements with G20 and BRICS+ member states.
AI based R&D will be financed in the amount of R$ 1.3 billion (RUB 18.6 billion) from which R$ 100 million (RUB 1.4 billion) will be spent to strengthen 184 AI centers network operating in universities, research organisations and enterprises. Artificial Intelligence Institute established under National Laboratory for Scientific Computing (LNCC) in 2023 will manage the network.
CNPq grants are envisaged for AI projects carried out at universities and research centers which enable to cover materials and equipment costs R$ 30-200 thousand (RUB 0.4-2.9 million) over two years. In total, R$ 753 million (RUB 10.8 billion) are planned for this item until 2028. Around R$ 100 million (RUB 1.4 billion) are allotted to interdisciplinary research using AI, and equal amount for joint projects with countries in Latin America and Africa.
For implementation of 37 AI-based technologies previously developed at Brazil universities and research centers R$ 435 million (RUB 6.2 billion) is available, including 37% from FNDCT and 63% from the federal budget. Funds are distributed among the following areas of AI application: healthcare — 70%, agriculture — 22.1%, education — 7.7%, and social development — 0.2%.
To stimulate Brazilian industry research in AI FINEP supports 500 projects of total R$ 9.1 billion (RUB 130.1 billion) until 2028 in 6 technological priorities identified under “New Industry Brazil” programme (see reference). FINEP grants are available up to four years and depend on size of company: up to R$ 3 million (RUB 42.9 million) for micro enterprise, up to R$ 5 million (RUB 71.5 million) for small enterprise, and up to R$ 8 million (RUB 114.4 million) for medium enterprise.
Reference:
"New Industry Brazil" programme launched by the Ministry of Development, Industry, Trade and Services with participation of National Council for Industrial Development in 2024 aims to increase export potential high-tech industries and accelerate the country’s industrialization through technological modernization, decarbonization and integration in to global value chains. The programme sets the following targets by 2030: 25% increasing the share of high-tech and medium-tech industries in Brazil’s total industrial production; cut carbon dioxide emissions by 30% per value added of industry's Gross Domestic Product (GDP). Technology priorities include: advanced and sustainable agriculture (organic fertilizers, agricultural technology, specialized software); healthcare (medicines, medical equipment); transport (electric vehicles, air transport, smart infrastructure); digitalization of industry (semiconductors, software, AI, Internet of Things); green energy (biofluels, hydrogen, solar panels); strategic technologies (radar, nuclear energy, space technology).
Source: ISSEK HSE according to the Ministry of Development, Industry, Trade and Services of Brazil.
Any size companies planning to develop and/or implement AI-based solutions in their business processes can receive a loan of up to R$ 15 million (RUB 214.4 million) to purchase relevant equipment for up to 48 months interest-free through the National Bank for Economic and Social Development (BNDES); annual interest rate of 8% applies after the grace period. Totally R$ 90 million (RUB 1.3 billion) is envisaged to support this instrument.
AI personnel training
Almost a fifth of support measures under the 2024–2028 plan aims human resources strength in AI and data analysis - R$ 1.2 billion (RUB 17.1 billion). Half of these funds is earmarked for AI courses development and implementation at all levels of education. Specifically R$ 183 million (RUB ~2.6 billion) are allocated for relevant higher education programmes. Additionally R$ 350 million (RUB 5 billion) from FNDCT are allocated for technical bachelor programmes, industrial master's degree and graduate school programmes contributing to a pool formation of practice-oriented specialists in AI. During BAIP realization, it is planned to train 5,000 specialists in AI development, including 50 AI semiconductor designers and 10 AI infrastructure experts, 15,000 specialists – end users of AI solutions in industry.
CNPq and CAPES allocate R$ 194.4 million (RUB 2.8 billion) to provide 650 master's and doctoral students scholarships studying AI-related programmes. R$ 152 million (RUB 2.2 billion) are to support 450 students in joint graduate schools at national and international universities. Standard postgraduate student scholarship in Brazil is monthly R$ 3.1 thousand (RUB 44.3 thousand) up to four years (if recipient does not work) and R$ 7 thousand (RUB 100.1 thousand) during staying abroad (see details «Господдержка аспирантов и постдоков в Бразилии»).
Government pays particular attention to strengthening ties between academic sector and industry. Specifically, R$ 150 million (RUB 2.1 billion) is being allocated through CNPq and FNDCT for AI projects that universities are implementing jointly with small and medium-sized enterprises (SMEs). A third of these funds are for 200 researcher fellowships hosting by SMEs. Such target fellowship overcomes standard one and is monthly R$ 17 thousand (RUB ~243 thousand) up to 30 months.
For the first time, the CNPq Research Productivity Programme has expanded its scholarship pool: approximately one thousand researchers employed in private sector and conducting research in AI will receive a monthly salary supplement of R$ 2,100 - 3,100 (RUB 30,000 - 44,300), renewable every two years depending on outcomes.
Since 2025 industry employees can improve their AI skills through specialized national platform offering 500 free courses (both in-person and on-line) jointly developed with country’s leading universities. 50,000 specialists will be trained through this platform by 2028.
Comment
Mikhail Gershman
Director of ISSEK HSE Centre for Science, Technology, Innovation and Information Policy
Brazil is consistently building a comprehensive national AI ecosystem that integrates research and computing infrastructure development, large-scale talent training and retention, as well as AI solutions implementation to industry, primarily at small and medium-sized enterprises. One of BAIP’s innovation is private sector involvement to public support system for science: for the first time AI researchers employed by commercial companies are eligible to receive prestigious "Research Productivity" scholarships. This measure, along with joint funding programmes for projects between universities and SMEs, aims at reducing the gap between science and industry and countering the “brain drain” from the country.
Brazil's experience in integrating AI into economic priorities is of particular interest to Russia. It is distinguished by a portfolio approach combining grant support, subsidized lending, creation of sovereign AI models in national language, and development of domestic computing infrastructure. Adapting such a model—with an emphasis on digital sovereignty, cross-sector collaboration, and a focus on socio-economic challenges—could strengthen the implementation of projects to ensure technological leadership and accelerate the digital transformation of industry.
Sources: Official websites of the Brazilian government; project results in accordance with the approved list of research and methodological support topics stipulated by the State Assignment of the HSE University.
The article is prepared by Brambila Martinez Francisco Javier
Lucas Adrian Garcia participated in collecting information
Предыдущий выпуск серии «Наука, технологии, инновации»: «Господдержка аспирантов и постдоков в Бразилии»
See also:
