Cases & Voices

Artificial intelligence is redefining how we research, teach, and collaborate—raising new questions about its impact and its human dimension. From Chile, Dr. Álvaro Soto, Director of the National Center for Artificial Intelligence (CENIA), reflects on the global and regional challenges of AI— topics he will address during his keynote session at TICAL2025, taking place in Costa Rica from November 11 to 13.

Created within the framework of Chile’s National Artificial Intelligence Policy (2021), CENIA promotes advanced research, technology transfer, and regional cooperation. Among its flagship initiatives are the Latin American Artificial Intelligence Index (ILIA)—whose third edition, launched in September 2025, includes data from 19 countries—and LATAM-GPT, a collaborative large language model developed by more than 50 institutions.

In conversation with the DeCLARA podcast, Soto discussed the main findings of the latest ILIA report and shared his vision of how Latin America can become a true actor—not just an observer—in the global AI revolution.

What brings us together as a region with regard to AI?
The central motivation is to encourage collaboration between countries. In Latin America, we share many common factors, such as similar challenges in areas like education and security, as well as a shared culture, history, roots and languages. This shared background makes it easier to create joint initiatives. Projects such as the Latin American Artificial Intelligence Index (ILIA) and LATAM-GPT emerge from this spirit, aiming to build shared knowledge and capacities. Our desire to work cooperatively and openly drives us, bringing together colleagues and institutions to collectively forge this path.

How can AI research be connected with its practical applications across different sectors?
Bridging the gap between research and implementation, or technology transfer, has always been challenging. However, we are currently experiencing a historic moment in which artificial intelligence has become truly disruptive and is facilitating dialogue with different sectors. Engaging industry or other fields from a scientific standpoint used to be difficult, but now the conversation flows more naturally.
At CENIA, we have been fortunate to be at the heart of this transformation at a time when communication between science and society is more important than ever. Our initiatives go beyond developing AI itself; they aim for sustainable, human-centred development, putting technology at the service of people.

How can artificial intelligence be considered an enabling infrastructure for science and development?
AI’s strength lies in its transversal nature. It’s hard to imagine a field in which it cannot have an impact, whether that be health, education, industry or scientific research. This makes it a unique technology that opens countless doors and possibilities.
The key lies in understanding its essential components. Why didn’t we talk about AI ten or twenty years ago with the same intensity as today? What has changed? Understanding these factors enables us to harness AI for the benefit of humanity.
Sometimes AI is adopted just to keep up with trends. While this may be an initial motivation, it’s crucial to go deeper and truly comprehend AI.
At TICAL2025, I will give a keynote speech addressing the core technical elements behind this revolution and how they can be translated into real opportunities. Unlike other disciplines, such as quantum mechanics, AI is something we can all relate to because intelligence is part of our everyday lives. In that sense, we are all, to some extent, experts in intelligence itself, which helps us to grasp the true magnitude and impact of this technology.

Nevertheless, there remains opposition to AI from some quarters and across various sectors.
It’s a disruptive technology and, if not managed properly, it can have negative consequences. These concerns are legitimate and must be taken seriously. We must keep technology under control and remember that it is a means to an end, not an end in itself. The true goal is to improve people’s lives — AI must serve humanity.
This requires an understanding of its scale, the amplification of its positive aspects and the avoidance of adverse effects. One of my main concerns is the disruption that AI may cause to the labour market. Today, we have what I call 'cognitive tractors'. Just as physical tractors once transformed agriculture by replacing manual tools such as hoes and rakes with machines that increased efficiency, we are now seeing a similar shift in the realm of cognitive tasks.
Many of these tasks will be affected, and the key question is how to ensure that these changes benefit society as a whole, rather than just a select few. These are crucial discussions to be having today. AI has already moved beyond the scientific sphere and is now a technology that we must study and understand in terms of its social impact.

What progress or setbacks has the region shown according to the latest ILIA?
Created four years ago, the Latin American Artificial Intelligence Index (ILIA) provides reliable data to inform decisions on AI development in the region. It evaluates three main dimensions:
First, enabling factors (infrastructure, connectivity, and talent). Second, development, which includes research, entrepreneurship and technology transfer, is an integral part of our mission. And third, governance (regulation and incentives).
Initially covering 12 countries, the project now includes 19 nations across Latin America and the Caribbean. Each country has its own national profile and comparative data. The most rewarding aspect has been observing the Index evolve into a regional reference tool, utilised by governments and institutions to develop public policies and inform their AI strategies.
The Index is also a truly collaborative effort, made possible by organisations across the region that contribute to and help gather data from diverse national contexts.

Which key findings stand out in the latest edition?
There are encouraging signs, but challenges remain.
On a positive note, several countries that were previously lagging behind have made significant investments. The Dominican Republic is a standout example of improvements in connectivity and infrastructure, while Costa Rica has strengthened its training programmes and human capital development.
There has also been growth in AI-based start-ups, and more countries are prioritising national AI strategies. All of this demonstrates increasing attention, motivation and tangible progress across the region.
However, there is still a lack of urgency. While some national strategies have been created, they lack the funding required to drive initiatives rapidly. Meanwhile, the pace of AI growth in regions such as Europe, Asia and North America is overwhelming. Rather than narrowing, the gap between us and those regions is actually widening.
We must act decisively and collectively to embrace the artificial intelligence revolution and become active participants, not just spectators, of this transformative era.

What role does international cooperation play in speeding up policies and capabilities?
That’s the foundation. In ILIA 2025, we analysed how different regions have adopted general strategies to face the AI revolution, together with Peruvian researcher Omar Flores. In the United States, the focus is on protecting and incentivising major tech companies such as OpenAI, Google, Meta and Amazon through policies that reinforce global leadership.
China takes a similar approach, combining state protection with extensive access to data and some degree of open-source collaboration, in an attempt to reduce the gap with North America. In Europe, there is a strong emphasis on AI factories and regulation.
The question is: what is Latin America’s AI model? I believe that collaboration and open sharing should be our approach. The recent evolution of AI shows that openness is a successful strategy.
Over the past two decades, AI’s explosive growth has largely been driven by democratised access to technology, data and open resources. Many of the most widely used AI frameworks, such as PyTorch and TensorFlow, and datasets were publicly available and were created by companies such as OpenAI, Google and Meta, and further advanced by academia.
Open collaboration and knowledge sharing are virtuous in themselves. Latin America, with its unique identity and shared values, could adopt this approach as a model for development and set an example to the rest of the world by showing how openness and cooperation can drive inclusive technological progress.

You are also collaborating on a GPT model specific to Latin America now. Can you tell us more about it?
Yes, that’s a concrete example. LATAM-GPT is an advanced language model that we are developing using data and resources from across the region.
The idea is to demonstrate that a collaborative, cooperative model can work for Latin America. Data comes from 50 to 100 institutions throughout the region. The project also includes a shared computing network because these large models require significant computational capacity, which no single country can provide on its own.
However, by working together, we can develop a regional AI infrastructure that can support large-scale models. Furthermore, sharing the resulting knowledge provides us with a collective advantage. 'Together we are stronger' is not just a phrase; it's a development model in action. LATAM-GPT aims to demonstrate that cooperation can be a successful path for Latin America.

TICAL 2025 has invited you to be a keynote speaker. What can participants expect?
I’m delighted to share CENIA’s experience and my perspective on the key elements behind the AI revolution, and explain how understanding these elements can help us to foster virtuous development. By 'virtuous', I mean artificial intelligence that enhances, improves and empowers us as a society, always serving people.

What is your vision for Latin America's place on the global AI stage in the coming decade?
Personally, I’m not particularly concerned about whether we become a global superpower. What I hope is that the region becomes happier and more prosperous thanks to AI. If AI helps everyone living in this part of the world to be more fulfilled and prosperous — not just a select few, but everyone, through inclusive technological progress — then the story of AI in Latin America will be a happy and successful one.