As artificial intelligence continues to evolve rapidly, governments across Latin America and the Caribbean face an urgent challenge: how to build the institutional capacity needed to govern a technology that is constantly changing.
Several countries in the region are already moving forward with AI strategies, laws, and regulatory proposals. El Salvador and Peru have existing AI laws, while Brazil, Chile, Colombia, and others are working on new regulatory frameworks. The Inter-American Development Bank has also proposed an enabling approach to AI regulation that seeks to balance innovation, rights, and responsible technology development.
However, passing laws is only one part of the challenge. The larger task is operational. Governments must be able to supervise AI systems, coordinate with technical experts, work with the private sector and academia, and update rules as technologies change. AI governance is not only about writing regulations; it is about building institutions that can implement, adapt, and respond.
For Latin America and the Caribbean, Spain offers a useful example of how legislative intent can be turned into an operational institutional ecosystem. Spain’s experience shows that governments do not need to wait until every legal detail is settled before they begin building the capacity to govern AI.
Spain began treating AI governance as a state policy linked to wider economic modernization. In 2020, the country created the Secretariat of State for Digitalization and Artificial Intelligence and published its National Artificial Intelligence Strategy. The strategy was supported by significant investment from the post-pandemic Recovery, Transformation, and Resilience Plan.
Over time, Spain’s governance structure continued to evolve. In 2023, AI governance became part of an independent Ministry for Digital Transformation and Civil Service. Later restructuring created separate responsibilities for artificial intelligence, digital services, connectivity, the digital economy, and data governance.
For countries in Latin America and the Caribbean, the exact organizational structure is less important than the broader lesson. Spain built a system around AI governance that included public agencies, investment bodies, advisory councils, foresight mechanisms, and multisectoral forums. This allowed the State to act, not only regulate.
This approach matters because effective AI governance requires more than legal authority. Governments need institutions with budgets, mandates, technical knowledge, and the ability to work across sectors. Without these capacities, AI strategies may remain on paper and fail to influence real-world development.
One of Spain’s most important steps was the creation of the Spanish Agency for the Supervision of Artificial Intelligence. Established in 2023 and operational in 2024, the agency became the first dedicated AI supervisory body in a European Union Member State. It was created before the European AI Act was finalized, showing that institutional preparation can begin before the law formally requires it.
The agency also runs Europe’s first AI regulatory sandbox. This testing environment allows companies to operate real AI systems under regulatory supervision, helping both innovators and authorities understand how rules work in practice. For emerging AI governance systems, sandboxes can provide a practical way to learn, test, and adjust before regulations are fully implemented.
Spain’s experience also shows the importance of flexibility. Its first national AI strategy was written in 2020, before generative AI became widely visible. By 2024, the country had to update its strategy to address new tools and priorities, including language models and a major investment package. In a fast-moving field, the ability to revise policy is just as important as the original strategy.
For Latin America and the Caribbean, this lesson is especially relevant. Countries in the region have different levels of institutional readiness, legal development, technical capacity, and public-sector resources. A single model cannot fit every country. What matters is building institutions that can learn, adapt, and coordinate as AI technologies evolve.
Governments can begin by creating or designating a lead AI authority with a clear mandate, budget, and coordination power. This authority does not need to be perfect from the beginning. It can grow over time, but it must have enough responsibility and visibility to bring together ministries, regulators, technical experts, private sector actors, civil society, and academia.
Countries also need to build the capacity to act, not only the capacity to write rules. This means creating public entities that can finance projects, support implementation, run pilot programmes, manage technical services, and guide responsible innovation. Without this executive layer, laws and strategies may lack the practical force needed to shape AI development.
Talent is another central issue. Skilled professionals in AI, data governance, cybersecurity, ethics, and digital regulation are in high demand. Governments must find ways to attract and retain this expertise through flexible hiring, competitive career paths, and opportunities for technical growth. Without skilled people inside institutions, AI governance bodies may struggle to keep pace with the technologies they are expected to oversee.
AI governance systems also need room to revise and improve. Periodic policy reviews, flexible mandates, regulatory sandboxes, and structured engagement with industry and civil society can help governments adapt as they learn. This is particularly important because AI risks and opportunities will continue to change as new systems, applications, and business models emerge.
For the region, Spain’s example is not a template to copy directly. Instead, it is a practical case showing that AI institutions can be built alongside laws, rather than only after laws are passed. This approach allows governments to begin developing expertise, testing regulatory tools, and creating coordination mechanisms while broader legal frameworks are still taking shape.
The key message for Latin America and the Caribbean is that AI governance is a long-term institutional process. It requires leadership, technical capacity, public investment, legal flexibility, and collaboration across sectors. Strong laws are important, but they must be supported by institutions capable of enforcing, updating, and applying them in practice.
As AI becomes more deeply embedded in public services, business, education, security, and everyday life, governments must prepare themselves to govern both current and future technologies. Building institutional capacity early can help countries protect rights, support innovation, reduce risks, and ensure that AI development serves public goals.
For Latin America and the Caribbean, the path forward is not only about regulating artificial intelligence. It is about strengthening the State for the digital era. Countries that build flexible, capable, and forward-looking AI institutions today will be better positioned to manage the opportunities and challenges of tomorrow.
