AI Talks

---

A Diverse Regional Landscape

The report groups countries into three categories based on their AI maturity:

• Pioneers: Chile (70.56), Brazil (67.39), and Uruguay (62.32) lead the region. They have strong infrastructure, specialized talent, and clear governance frameworks.
• Adopters: Countries such as Mexico, Argentina, and Colombia show progress but still face gaps in research capacity and investment.
• Explorers: Scoring below 35 points, nations like Bolivia, Venezuela, and Paraguay are just starting to build the foundations for AI development.

Encouragingly, several “late adopters”—including Ecuador, Costa Rica, Guatemala, and the Dominican Republic—are accelerating fast. Their investments in connectivity, education, and national AI plans show that progress is possible even without a large industrial base.

Where the Momentum Lies: Generative AI and Open Source

Two forces are rapidly reshaping Latin America’s AI ecosystem: Generative AI and Open Source software.

Generative AI—popularized by tools like ChatGPT and Gemini—has spread faster in Latin America than almost anywhere else. The region now accounts for 15–20% of all global downloads of AI applications. Thanks to easy-to-use interfaces and low barriers to entry, millions of users across Chile, Costa Rica, Peru, and Panama are adopting these technologies in daily life and business. In just a few years, AI has gone from an elite research topic to a common digital tool.

Open Source AI offers another advantage. By sharing code and models freely, smaller nations can innovate without depending on costly foreign licenses or infrastructure. Countries such as Honduras, El Salvador, and Cuba are using this model to build local solutions and foster collaboration. A standout example is LatamGPT, the first large open-source language model developed for and by Latin America. It aims to reflect the region’s cultural and linguistic diversity—including Spanish, Portuguese, and Indigenous languages—while promoting digital sovereignty.

The Paradox: Big Ambitions, Limited Action

Despite this momentum, the ILIA exposes a deep paradox: great enthusiasm, limited execution.

Latin America represents 6.6% of global GDP, but attracts only 1.12% of worldwide AI investment. Average investment relative to GDP per capita is six times lower than the global benchmark. Brazil and Costa Rica are beginning to reverse the trend, yet no country has crossed the world average.

Research and innovation remain highly concentrated. Brazil and Mexico host 68% of the region’s AI researchers, while just five countries—Brazil, Mexico, Chile, Colombia, and Argentina—produce 90% of its scientific publications. This unevenness reduces visibility in international forums and limits regional knowledge exchange.

Infrastructure shows similar imbalances. Brazil alone holds over 90% of Latin America’s high-performance computing capacity. More than half the countries lack this essential resource, which constrains their ability to train or deploy advanced AI models.

Then there is the data gap. The region generates massive amounts of information but struggles to make it openly available or standardized. Without accessible and interoperable data, it is difficult to build precise, locally relevant AI applications—whether for public health, agriculture, or climate resilience.

The Talent Challenge

Education is another bottleneck. Basic AI literacy is growing fast, but advanced training is scarce. Thirteen of the 19 countries studied do not teach AI skills in school curricula, and 11 lack doctoral programs in the field. As a result, Latin America produces few AI specialists capable of developing cutting-edge systems. Costa Rica stands out as an exception, with a strong focus on professional AI skills.

Without sustained investment in specialized education and research, the region risks deepening its dependency on imported technology rather than building its own capacity.

The Governance and Sustainability Gap

Nine Latin American countries have adopted national AI strategies—a promising step—but most lack budgets, implementation plans, or measurable indicators. The report summarizes this as “much plan and little action.”

Equally concerning is the lack of attention to environmental sustainability. Data centers and model training consume enormous amounts of energy, yet few national strategies account for their carbon footprint. Integrating sustainability into AI policy could align technological progress with the region’s climate goals.

AI also offers underused potential for digital democracy. Instead of using AI to enhance citizen participation, transparency, or accountability, most public initiatives are limited to informational chatbots.

From Consumers to Creators

The message from ILIA 2025 is clear: Latin America stands at a crossroads. The region has the creativity, connectivity, and growing user base to thrive in the AI era. But to transform enthusiasm into tangible progress, three priorities stand out:

1. Invest in Talent and Infrastructure. Expand advanced education programs, strengthen university-industry partnerships, and build shared regional computing facilities.
2. Adopt Open and Sustainable Practices. Promote open data, open science, and open-source AI development while integrating environmental criteria into every project.
3. Turn Plans into Policy. Move beyond strategy papers to concrete budgets, regulations, and measurable goals.

AI progress doesn’t depend on a massive industrial economy. What matters most is coordination in education and infrastructure, inclusion in data and research, and a shared regional vision for policy and innovation. With collaboration—through initiatives like LatamGPT and strong public-private partnerships—Latin America can move from being a consumer of imported technologies to a creator of its own digital future.

Conclusion

The Index of Artificial Intelligence in Latin America 2025 captures a region in motion—enthusiastic, uneven, and full of potential. The technological gap remains wide, but a way forward is visible.

Latin America’s future lies in turning plans into action: investing in people, promoting open and sustainable innovation, and strengthening collaboration across borders. With consistent effort, the region can transform the current AI wave into a lasting engine of competitiveness, creativity, and inclusion

---

Reference

Durán, R., Moreno, A., Adasme, S., Rovira, S., Jordán, V. y Poveda, L. (Coords.) (2025). Índice Latinoamericano de Inteligencia Artificial (ILIA) 2025. Project Documents (LC/TS.2025/68). Economic Commission for Latin America and the Caribbean and National Artificial Intelligence Center. United Nations publication. https://indicelatam.cl/

This post was researched and written with the assistance of various AI-based tools.