We are living in a critical and exceptional moment. Artificial intelligence (AI) has ceased to be a distant promise or a discussion reserved for specialists: it is already redefining productivity, power, institutions, and everyday life at an unprecedented speed. For Latin America and the Caribbean, the decisive question is not whether AI will be adopted, but how, under what conditions, and in service of what development project. It can open the greatest opportunity for well-being in generations, but it can also deepen dependencies, inequalities, and risks. The challenge is not only to digitize more, but to build our own capacities so that AI transforms the economy, the state, and people’s lives with rules, priorities, and values also defined from the region.
We are living through more than a time of change: we are witnessing a change of era. The steam engine, electricity, and the internet altered the trajectory of humanity, but none spread as quickly as today. Electricity took more than four decades to achieve mass penetration in homes; ChatGPT reached one hundred million users in just two months. Nearly 2.4 billion people use generative AI tools every month, while the computing power to train cutting-edge models doubles approximately every five months. This acceleration demands that innovation be paced with our ability to understand, supervise, and govern it.
The challenge is especially urgent for a region simultaneously facing green, technological, and demographic transitions. AI intersects with and conditions the others: the climate response, protection of our environmental assets, consolidation of clean energy matrices, productivity increases, and closing gaps will increasingly depend on quality data, computing, connectivity, specialized talent, and models tailored to our realities. All this occurs in a context of fiscal tightness, high debt, informality, persistent poverty, and low growth. Precisely for this reason, AI must help us do more and better: raise productivity, improve spending, and expand the impact of public policies.
This transformation does not happen in a geopolitical vacuum. Data, semiconductors, computing, and AI models are already strategic assets. In this new geography of power, digital sovereignty should not be understood as isolation or autarky, but as real decision-making capacity: governing our data, protecting people, having adequate regulatory frameworks, adapting technology, developing our own capacities, and participating in strategic links of the value chain. The greatest risk would be resigning ourselves to passively consuming technologies created by others and accepting rules in whose definition we had no voice.
No country in the region can influence global AI rules alone. A coordinated Latin America and the Caribbean, however, can share capacities, turn local solutions into regional goods, and negotiate from a stronger position. Decisions about security, access to computing, data governance, standards, employment, intellectual property, foundational models, and limits on autonomous systems cannot be reserved for a handful of companies, powers, or forums where the region is barely represented. With more than 650 million inhabitants, cultural diversity, and its own development challenges, Latin America and the Caribbean must participate with voice, proposals, and technical capacity. That cooperation must translate into a practical agenda: modern regulation, data protection, competent institutions, resilient infrastructure, talent training, access to computing, and regional coordination mechanisms. It is not about choosing between opening up to the world or building our own capacities, but about participating from a stronger base and contributing to AI in the service of our peoples.
The region has the conditions to build that path. It possesses critical minerals — lithium, copper, rare earths, and other essential inputs — and one of the greatest renewable generation potentials on the planet. The challenge is not to repeat the history of exporting wealth without adding value: turning that endowment into industry, knowledge, and productive linkages associated with the AI economy. Its expansion consumes enormous amounts of electricity and faces growing energy constraints. Our clean energy can sustain computing capacity installed in the region, indispensable for training, adapting, and deploying solutions with less external dependence.
We also have a less visible but strategic heritage: our languages, content, cultures, and ways of understanding the world. If we want AI to understand our reality and address our gaps, we need representative data and technological capacity to use it responsibly. Otherwise, systems trained mainly with foreign data and perspectives can reproduce biases, ignore contexts, and homogenize regional diversity. Digital sovereignty is, in this sense, also cultural sovereignty.
This perspective should not confuse prudence with immobility. AI opens extraordinary possibilities: new scientific discoveries, more personalized education, energy efficiency, and a productivity leap capable of improving well-being. It can put tools previously reserved for large corporations into the hands of a small company; transform agriculture, modernize industry, strengthen public services, and accelerate innovation. For a region that has dragged decades of low productivity growth, giving up that opportunity would be irresponsible.
It would be equally irresponsible to ignore the risks. Job displacement, environmental pressure, misinformation, and difficulty distinguishing truth from falsehood can weaken social cohesion and democratic trust. Recent history warns us: during the 2008 financial crisis, the region did not grasp the massification of smartphones nor the impact social networks would have on education, democracy, and human relationships. It took us about 15 years to understand part of those consequences. We cannot afford to be distracted again.
These concerns no longer come only from academics, regulators, or critical voices in the industry. Those at the technological frontier also raise the tone. Dario Amodei warns of rapid and painful labor disruption, risks of more autonomous systems, malicious uses, and extreme concentration of power. Sam Altman insists on safeguards, independent evaluations, and international coordination. Bill Gates poses an eloquent dilemma: AI can be the greatest equalizer of our time or a new source of injustice if its impacts are not anticipated. From different perspectives, they all converge on something essential: innovation advances faster than the response capacity of our institutions.
Pope Francis, addressing the G7, also defined AI as a tool as fascinating as it is fearsome and recalled a decisive distinction: a machine can make algorithmic choices, but deciding — weighing consequences, exercising prudence, and assuming responsibility — corresponds to the person. That is why he defended that there should always be a significant space for human control. More recently, Pope Leo XIV, in continuity with the humanist tradition expressed in Magnifica Humanitas, has reiterated that the benefits and risks of AI must be judged from human dignity, the common good, and respect for the cultural and spiritual diversity of peoples. That moral guidance is indispensable: technology cannot become an end in itself nor displace the center of social life; it must always remain at the service of people, their freedom, dignity, and integral development.
The response requires avoiding two extremes: naive enthusiasm and paralyzing pessimism. AI alone does not produce prosperity. Great productivity leaps appear when innovation combines with talent, effective institutions, reliable data, and new ways of organizing work. Adopting it does not mean buying tools or automating processes without criteria; it means transforming capacities, decisions, and responsibilities. Technology opens possibilities; institutions and collective action determine whether they become shared well-being.
That is why CAF — the development bank of Latin America and the Caribbean — has placed AI and its enabling capabilities at the center of its development mission. In April of this year, we created the Digital Transformation Management and incorporated it among the institutional strategy’s mission objectives for the next five years, with the mandate to help countries translate AI and digital technologies into concrete results: greater productivity, better public services, more reliable institutions, and new opportunities for people and companies.
That commitment requires resources on the scale of the challenge. We will double the annual funding allocated to initiatives with significant AI and enabling digital capabilities components: from an average close to 1.2 billion dollars to around 2.4 billion. This will allow surpassing 12 billion dollars in the next five years and will make AI, data, computing, connectivity, and institutional transformation centrally present in more than 30% of the initiatives financed by CAF. But resources are not enough: it will be essential to bring in other multilateral organizations, financial institutions, investors, companies, academia, and civil society around a shared vision.
The agenda must act on three simultaneous fronts: strengthen public capacities to use AI safely, responsibly, and effectively; finance connectivity, data centers, processing infrastructure, cybersecurity, and other critical enablers; and accompany productive transformation, especially of micro, small, and medium enterprises, so that AI adoption translates into innovation, better processes, and quality jobs.
The fundamental question is not what technology can do, but what society we want to build with it. Latin America and the Caribbean must decide whether it will be a spectator or a protagonist: whether it will consume what others create or leverage its minerals, clean energy, talent, data, and cultural diversity to participate with its own voice in the rules of this new era. That decision requires political dialogue, technical cooperation, and a regional agenda capable of turning national capacities into shared goods.
From CAF, the common home of the countries of Latin America and the Caribbean, we invite building that dialogue and transforming this agenda into concrete cooperation. This challenge cannot be addressed in a fragmented way: it requires governments, the private sector, academia, civil society, and international partners working together, with a shared and deeply regional vision. Let us not fall prey to naivety, pessimism, or inaction. Let us move forward even amid questions. In integration, knowledge, and coordinated action lies our capacity to make artificial intelligence a tool at the service of people and to write, together, our own page in this transformation.
