Brazil’s public sector is preparing to debate not only how artificial intelligence can be used, but also who should control the systems behind it. The issue will be at the center of the II Serpro AI Week, scheduled for November 4 to 6 in Brasília, with online broadcasting and free participation.
Announced by Brazil’s federal data processing company on October 9, the event will bring together representatives of government, universities, research centers and technology companies. Its theme, “Sovereign AI: Data, Technology and Decision-Making Power,” places institutional control at the heart of discussions about artificial intelligence in public services.
The event is relevant because AI systems are increasingly connected to processes that organize information, support public administration and influence decisions. In this context, the debate extends beyond the creation of digital tools. It also involves the origin of the data, the models selected for specific tasks and the computing infrastructure used to process information.
Control goes beyond the visible AI tool
For public institutions, choosing an artificial intelligence solution can involve a chain of technical and strategic decisions. Data must be managed, models must be evaluated and computing resources must be available to run the systems. These elements can affect how much control an institution retains over technology used in government operations.
The Serpro event is therefore framed around a broader question: what does the public sector need to master before relying on artificial intelligence? The answer suggested by the agenda includes more than access to an application. It includes the ability to understand and govern the foundations that make the application work.
This perspective also brings issues such as security, transparency and dependence on private providers into the discussion. The event announcement does not present a single solution to those challenges, but proposes a debate on how public institutions can make decisions about AI while maintaining control over strategic parts of the technology.
Why data, models and computing capacity matter
Data is one of the central elements in any AI system because it influences the information available for processing and analysis. The choice of a model is also strategic: different models can have different technical characteristics, uses and operational requirements. Infrastructure, in turn, determines where and how information is processed.
By grouping these three areas in its theme, Serpro signals that artificial intelligence governance cannot be reduced to software procurement. Public-sector organizations also need to assess the technological conditions that support their systems and the consequences of depending on infrastructure controlled by outside providers.
That discussion has a direct institutional dimension. When technology is used in public services or decision-making processes, questions about accountability and transparency become part of the deployment process. The greater the role of AI in government operations, the more important it becomes to identify who manages the data, selects the models and maintains the computing environment.
Government, research and industry in the same debate
The II Serpro AI Week will include participants from the public sector, academia, research institutions and technology companies. This combination reflects the range of expertise required to build and operate artificial intelligence systems: government contributes the public-service perspective, while universities and research centers expand scientific knowledge and companies bring technological capabilities.
The event’s agenda also points to cooperation among government, science, businesses and research networks as a path toward national AI infrastructure. The focus is not limited to a specific agency or product. Instead, it addresses the conditions needed for Brazil’s public institutions to participate in decisions about the technologies they use.
According to the Serpro announcement, the event will take place in Brasília from November 4 to 6 and will be available online. The format is intended to expand access to a discussion that concerns both technology specialists and the broader public affected by digital government services.
A wider national technology context
The debate is also connected to the federal government’s broader technology priorities. The 2026 Message to the National Congress includes planned investments in areas such as artificial intelligence, quantum technologies, advanced communications and researcher training.
The document provides context for why infrastructure and skills are part of the conversation. Developing AI capacity requires more than adopting ready-made systems; it also depends on research, technical expertise and the networks that support innovation. For the public sector, these factors are linked to the ability to evaluate technologies and make informed decisions about their use.
With registration open and participation offered both in person and online, Serpro’s event will serve as a forum for examining how Brazil can approach artificial intelligence as an issue of governance and technological capacity. The central question is not simply whether government should use AI, but how it can retain knowledge and decision-making power over the systems that support public administration.
