Portuguese 
English 
Spanish 
5 min of reading 
0 comments 
Critical situation faces countdown, as inertia persists for the approval of the project regulating the sector (Bill No. 2,338/2023), still ‘stuck’ in the National Congress, due to ‘endless discussions’ and nothing proactive. Another determining factor is the lack of computational infrastructure. Meanwhile, the ‘brain drain’ to other markets with greater legal security is increasing.
Brazil may be falling behind (and irreversibly) in the global race for data centers and Artificial Intelligence (AI) if it does not reverse, in the short term, the current sluggishness in the process of sector regulation (Bill No. 2,338/2023, ‘stuck’ in the National Congress). The consequence is the delay in investments in high-performance businesses, which is expected to compromise computational infrastructure.
The grim forecast comes from the director of Nvidia’s Enterprise Division – the world’s most valuable company by market value and absolute leader in the manufacture of semiconductors (microchips) and infrastructure dedicated to Artificial Intelligence (AI) – Márcio Aguiar, emphasizing that without robust local projects in the field of data centers, the trend is the exodus of data scientists, ‘machine learning’ engineers, and researchers to opportunities, including to neighboring competitors, in Latin America itself.
-
Brazil is generating so much clean energy that the ONS had to implement an unprecedented emergency plan to contain the solar and wind surplus in the grid.
-
Electric scooters have helped reduce cars on the streets, but their batteries also age: Lime has made an agreement with Redwood to recycle used packs in the USA, Germany, and the Netherlands and turn tech waste into raw material for new electric vehicles.
-
While many drivers still fear running out of charge on the road, CATL targets lithium-air batteries with a theoretical density close to gasoline and rekindles the promise of electric cars with a range of over 1,600 km.
-
Researchers created turbulence in the laboratory, reversed the energy flow within the fluid, and opened a new avenue for understanding ocean currents, climate, and sea pollution.
https://marcellosigwalt.substack.com/p/atraso-em-adaptacao-a-novas-tecnologias
Official initiatives ‘dive’ into ‘endless’ discussions
Meanwhile, the Brazilian Artificial Intelligence Plan (PBIA) and the Redata program (Special Incentive Regime for Cloud Computing and Data Centers) are ‘immersed’ in endless discussions in Brasília.
The PBIA foresees investments of R$ 23 billion, focusing on infrastructure, training, and priority areas such as health and education. However, for it to be effective, the initiative requires faster government execution and efficient integration between the public sector, academia, and the private sector.
After being approved by the Chamber of Deputies, the Provisional Measure establishing Redata ended up ‘lapsing’ (expired) in the Senate. It provided for the suspension of federal taxes for up to five years, aiming at the acquisition of machines, equipment, and technology components, with companies committing to productive investments focused on sustainability and expanding the installed technological capacity in the country.
Inertia may condemn the country to merely ‘consumer of ready-made solutions‘
If the ‘slowness’ in decision-making in Brazil becomes entrenched, Aguiar admits that the country risks perpetuating its ‘atavistic’ dependence on foreign technology, limiting itself to being a mere consumer of ready-made solutions, instead of producing cutting-edge technology. This structural pattern, which repeats across Latin America, is a limiting factor in the creation of local intellectual property and keeps companies dependent on the volatility of costs determined by global providers.
Echoing Aguiar’s forecasts, Softbank partner Eduardo Vieira, participating in the BM&C News Brazilian Week 2026 program, got straight to the point: “I think we have a very large mismatch between what is happening internationally and what could be happening in Brazil and is not happening.”
Highlighting that abroad, currently, the market has been emphasizing “investments in AI infrastructure layers, data centers, energy, chips, silicon, not to mention the foundation for companies and the new economy that will emerge, with AI-based applications in daily life, and transforming the industry, as was the case with the iPhone and Amazon, among others.”
Commenting that investments in this cutting-edge technology “are very concentrated in the United States, Vieira believes that Brazil is missing this opportunity, despite having a clean energy matrix and technology to attract data centers, but has not been able to attract foreign capital to this area.” As an example of market loss, the Softbank partner recalls that OpenAI is investing in a data center in Argentina, while Paraguay is advancing in the same direction.
Solution begins with a ‘clear’ regulatory policy
To reverse the negative trend, Vieira believes that the country urgently needs to offer “a very clear regulatory policy, a national project, so that we do not fall behind in this technological race, where the US and China are far ahead, with investments in the trillions of dollars. Although optimistic, stating that ‘there is still time’ for Brazil to keep up with the frenetic pace of AI on the planet, Vieira warns: “If there is political will, a convergence of interests, and if the government understands what is happening abroad and prioritizes this dialogue with the private sector, there is still time, but we do not have much time, as we are reaching a critical point that may soon be beyond reach.”
Data from the Ministry of Mines and Energy (MME) reveal that connection requests for new data center projects grew by 330% from 2024 to 2025. At the end of last year, the study already accounted for 28.5 GW of demand requested for projects planned until 2038. Currently, the 205 existing data centers in the country, in phases ranging from operation to projects under construction, total R$ 114.5 billion in estimated investments.
In an interview with Bloomberg Línea, the Nvidia executive does not mince words: “We are missing the opportunity to board this passing train,” he says, adding that “technology is evolving so quickly that every one or two months of delay already puts us in a new era of artificial intelligence.”
Regarding the Brazilian inertia on regulation, Aguiar reveals that smaller economies than Brazil and Mexico have shown ‘greater regulatory agility to attract investments and retain human capital’, especially Chile, Uruguay, El Salvador, Costa Rica, Ecuador, Dominican Republic, and Colombia.
Challenges of the Brazilian AI ecosystem
Infrastructure and hardware: The high cost of advanced equipment, such as graphics processors (GPUs), makes cutting-edge research and the training of large language models (LLMs) much more expensive and difficult for national companies and research centers.
Legislation and regulation: The delay in approving regulatory frameworks (such as Bill No. 2,338/2023, which is under consideration in Congress) creates legal uncertainty for investors and companies wishing to develop local solutions.
Inequality and adoption: The use of AI is very unequal in the country. It is concentrated in the South and Southeast regions and among the higher social classes, which prevents the democratization of productivity gains provided by the technology.
Be the first to react!