Saturday, July 20, 2024
- Advertisement -

    Latest Posts

    Views: What is AI Sovereignty and Why Brazil Can Lead the Global Debate About It?

    By Luca Belli

    Recently, in Rio de Janeiro, a top executive from Nvidia, a leading AI computing company, solemnly affirmed that “Brazil can still be AI sovereign.” Brazilians only need to become users of Nivida services and purchase Nvidia AI Graphic Processing Units (GPU), he added. While the former statement is correct, the latter seems misleading at best. 

    After the launch of our research on AI Sovereignty at the UN Internet Governance Forum, in October 2023, multiple AI companies have jumped on the AI sovereignty bandwagon, branding their products as the ideal solutions to become AI sovereign. It seems rather naïve to reduce AI sovereignty to the mere purchase of computational power from a foreign vendor. Moreover, a sceptical observer may consider slightly self-interested the equation of technological dependence from one company as being AI sovereign. 

    What is AI sovereignty?

    AI sovereignty is the latest evolution of the digital sovereignty debate and has enormous relevance for global geopolitics of technology. As I argue in recent research, it is essential to consider this type of sovereignty as the capacity to understand, develop and regulate AI systems. This conception stems from what I define as “good digital sovereignty,” a conception based on the fundamental right to self-determination, which rejects authoritarian views of sovereignty.

    This understanding is key to constructing a positive agenda for AI governance, based on the sustainable development and regulation of AI systems. It also looks increasingly palatable to an unlikely AI governance champion: Brazilian President Luiz Inacio Lula da Silva, known as Lula. 

    Speaking at the National Science and Technology Council, President Lula encouraged the promotion of a different conception of AI, shaped by and for the Global South and for the world. He intends to present such vision, at his opening of the UN General Assembly, in September.

    Importantly, Lula is one of the very few political leaders enjoying legitimacy and respect at the global level and having a clear vision of development. These are very rare features amongst current statesmen. If played well, his move can have remarkable consequences for global tech policy.

    From Brazil to the Global Majority

    Both the Lula administration and global policymakers are increasingly understanding that building AI (sovereignty) requires systemic thinking, to be fully aware of the “Key Enablers of AI Sovereignty”. To be digitally sovereign one needs solid research, development, and regulation of several enabling elements: data, algorithms, computational capacity, connectivity, energy, human resources, cybersecurity, and awareness of AI risks. These elements are all equally important and interlinked. 

    Focusing only on some of them with unorganised initiatives is pointless. The current Brazilian AI strategy, elaborated by the previous federal administration, tellingly proves this point. The document is incomplete, already obsolete, and contains no operational elements. It is a nice wish list, but not exactly ‘strategic’ for a strategy.

    This frustrating status quo, however, offers a golden opportunity to rebuild the Brazilian approach from scratch and to use it as a blueprint for the Global South. The federal government is redefining the national strategy and, if all the enablers are considered, the results can become an AI-sovereignty strategy model for the Global Majority and beyond.

    Clouds on the horizon?

    Paradoxically, the position of Nvidia, mentioned above is particularly useful to understand the state of technological dependence of the AI market. “Doing” AI is simply impossible without specialised semiconductors, GPUs, and access to cloud computing. 

    This latter market has increased by 20% in 2024 and most of it is concentrated in a few companies, namely Amazon, Microsoft, and Google. All of them depend on Nvidia’s equipments for AI. Consequently, they are all massively investing in R&D to escape dependence on Nvidia’s products and services. They clearly have a different conception of AI sovereignty.  

    It is not a coincidence that most AI startups willing to scale end up being acquired by or needing to partner with the companies mentioned above. Being dependent on someone, however, can hardly qualify as being sovereign. This is why SERPRO, the largest publicly owned IT enterprise in Brazil, created its own cloud computing platform. It openly states its goal is to “guarantee data sovereignty and privacy” for the Brazilian government.

    Indeed, the clouds in cloud computing are far from ethereal, abstract, and fluffy. They are probably the most physical part of AI systems. This also means that to power and cool such essential infrastructure, enormous amounts of energy and water are needed. It is not a coincidence that all cloud giants are striking deals with energy companies, including nuclear operators. OpenAI’s CEO is even boasting its investments in Helion, a startup exploring nuclear fusion.

    Brazil, on the other hand, is a leader in renewable energy. Few people know that the tropical giant is not only one of the few energy-independent countries in the world. It produces roughly 80% of its energy needs from renewables. Furthermore, besides holding the rotating presidency of the G20, Brazil will host the COP-30 and the BRICS summit in 2025, thus having a unique opportunity to conjugate the AI and environmental agendas. 

    Connecting global agendas

    This coincidence of multiple main-stage roles is an exceptional opportunity. If Brazil plays its cards well, it may succeed in strongly advancing and shaping discussions on environmental protection, equitable technological development, and fair trade in less than two years. None of these will be easy but Brazil has a unique blend of legitimate developmental aspirations, solid diplomacy, and key environmental role to achieve the task.

    AI sovereignty is a blend of socio-environmental justice, education, meaningful connectivity, data protection, and cybersecurity. Without considering all the enablers of AI sovereignty, neither Brazil nor any other country will ever enjoy self-determination. The tropical giant has an opportunity to demonstrate to the world that this is the essence of sovereignty: the fundamental right to define one’s technological, economic, and cultural evolution. 

    Moreover, Brazil can claim to have been a pioneer of digital sovereignty for more than two decades, even if this is largely unrecognised. The first Lula administration promoted Free and Open-source Software (FOSS) as state policy as far back as in 2003. FOSS is a unique strategy to avoid software dependency. This pioneering spirit can and should be replicated by Brazil and others, increasing cooperation and capacity building to promote the development of open AI models.

    Sovereign, not isolated

    Holding the rotating presidency of the G20 in 2024 and of the BRICS next year, Brazil has a unique opportunity to show that being sovereign does not mean being isolated. It means being willing and able to define one’s digital future, while cooperating with as many partners as possible, to avoid being technologically dependent. 

    In this context, learning from international best practices is vital, but copying and pasting existing approaches is useless. A good example is the debate on how to regulate AI-related risks. This issue is of paramount importance but Global South countries are understanding that copying and pasting the new EU AI Act is worthless if one is unable to understand how AI systems function and produce them. 

    We have already entered an era of AI-led transformation, impacting our economies, societies, democracies, and environment. We cannot revert this, but we still have time to upgrade our sovereignty, constructing a digital dimension of for our fundamental right to self-determination. 

    What is at stake is the capacity of countries and peoples to define their digital future and avoid being AI colonies. This should be the consistent message of Brazil to the UN, the G20, the BRICS and COP-30.

    Luca Belli is Professor at FGV Law School where he directs the Center for Technology and Society.

    Also Read:

    The post Views: What is AI Sovereignty and Why Brazil Can Lead the Global Debate About It? appeared first on MEDIANAMA.

    Latest Posts

    - Advertisement -

    Don't Miss

    Stay in touch

    To be updated with all the latest news, offers and special announcements.