The AI Race Has Two Superpowers. Where Is Latin America and the Caribbean?
As the United States and China pursue competing visions of artificial intelligence, Latin America and the Caribbean confront a question that receives considerably less international attention: will the region become another market for imported intelligence, or develop the infrastructure, institutions and productive capabilities necessary to participate in creating its economic value?
There is something rather peculiar about watching the world’s most powerful economies debate the future of artificial intelligence while much of the developing world appears to be waiting to discover what that future will look like. In the United States, technology executives, investors and policymakers discuss artificial general intelligence, scientific breakthroughs, national security and the extraordinary economic possibilities associated with increasingly capable machines. In China, technological development is similarly connected to national competitiveness, but discussions frequently emphasize industrial applications, commercial deployment and the integration of artificial intelligence into existing productive systems. Meanwhile, countries across Latin America and the Caribbean confront a rather different question: where exactly do we fit into all of this?
The question becomes especially important when one considers how our region has historically participated in major technological and industrial transformations. We have imported machinery, automobiles, telecommunications equipment, computers and countless other products developed elsewhere. We have adapted those technologies to our circumstances, sometimes with extraordinary ingenuity, while frequently remaining dependent upon foreign manufacturers, intellectual property owners and the financial institutions supporting their industries. Participation in the global economy has not always translated into equivalent participation in the ownership of the technologies shaping it.
Artificial intelligence introduces the possibility of another such transition, although its economic consequences may be considerably more extensive. Unlike technologies designed to perform a narrow physical function, AI systems can assist with research, software development, financial analysis, administration, engineering, communication and numerous other activities that influence how institutions operate. Their capabilities remain imperfect, but their potential applications extend across almost every productive sector. The question is whether our countries will approach this development primarily as consumers or recognize an opportunity to strengthen their own economic capabilities.
I recently watched an interview with researcher and writer Yi-Ling Liu examining the differences between American and Chinese attitudes toward artificial intelligence. Her observations prompted me to consider a perspective that is frequently missing from the international conversation. The future of AI is often presented as a contest between Washington and Beijing, with the rest of the world appearing primarily as territory over which technological influence will be exercised. But Latin America and the Caribbean are not merely markets waiting to be divided between competing technology companies.
We are societies with our own industries, institutions, economic vulnerabilities, cultural identities and development ambitions. The technologies we adopt should ultimately serve those circumstances rather than require us to organize our futures entirely around the priorities of larger countries.
The question is not simply who wins the artificial intelligence race. It is whether the rest of us understand that we are already participating in it, whether we intended to or not.
America Is Racing Toward Intelligence. China Is Thinking About What to Do With It.
One of the most interesting distinctions emerging from Liu’s interview concerns the language through which artificial intelligence is discussed. In American technology circles, considerable attention is devoted to the development of increasingly general AI systems capable of performing sophisticated intellectual tasks across multiple disciplines. The possibility of artificial general intelligence has become a central subject of investment, research and public debate, accompanied by questions concerning safety, employment, national security and the future relationship between humanity and machines.
China’s approach is often described through a different emphasis. Government initiatives and industrial strategies have placed substantial attention on integrating AI into manufacturing, logistics, transportation, healthcare and other existing economic activities. The objective is not exclusively to develop increasingly capable models, but also to improve the productivity and competitiveness of industries that already generate substantial economic value.
These distinctions should not be exaggerated. American companies are deeply involved in commercial AI deployment, while Chinese researchers and technology companies also pursue advanced models and frontier capabilities. Both countries recognize the strategic importance of computing infrastructure, semiconductors, scientific research and industrial applications. Their approaches contain considerable overlap, even where political institutions, market structures and public narratives differ.
Nevertheless, Liu’s observations illuminate an important economic distinction: developing intelligence and applying intelligence are related but separate activities.
A country may possess sophisticated AI research institutions without achieving corresponding improvements in productivity across its wider economy. Another may develop fewer frontier models while generating substantial benefits through the adoption of existing technologies in agriculture, manufacturing, public services and commerce. The economic value of artificial intelligence ultimately depends not only on what the technology can accomplish, but also on whether institutions possess the capacity to use it effectively.
For Latin America and the Caribbean, that distinction is consequential. Most of our countries do not possess the financial resources necessary to compete directly with the largest American and Chinese frontier AI laboratories. Training advanced models requires substantial computing infrastructure, specialized expertise, electricity and access to sophisticated semiconductor technologies. Attempting to reproduce those capabilities independently may be economically unjustifiable for many smaller states.
But the inability to compete at the frontier does not eliminate the possibility of meaningful technological development. Countries can develop specialized applications, improve institutional efficiency and establish competitive businesses using internationally available AI technologies. The relevant question is which capabilities should be developed domestically, which should be acquired through partnerships and which can be accessed commercially without creating unacceptable dependencies.
That is a considerably more useful starting point than assuming technological leadership must mean producing the world’s most powerful model.
The Difference Between Buying Intelligence and Owning Capability
There is an important distinction between purchasing access to artificial intelligence and developing an economy capable of creating value through it. A business can subscribe to an AI service, automate certain administrative activities and improve its operations without possessing any underlying intellectual property. Such arrangements may be commercially sensible, particularly when developing equivalent technology independently would be expensive or unnecessary. The difficulty emerges when an entire economy becomes dependent on imported systems without developing complementary domestic capabilities.
Consider how this relationship might develop across the Caribbean. Hotels purchase AI-enabled customer-service systems, financial institutions acquire automated compliance tools, governments procure digital administrative platforms and businesses integrate foreign software into their daily operations. These investments may improve productivity, reduce costs and expand access to useful services. But much of the associated revenue may flow toward overseas technology providers, particularly where the software, computing infrastructure and intellectual property remain externally owned.
This is not inherently different from other forms of international trade. Countries routinely purchase equipment and services that would be inefficient to produce domestically, and access to international technology can generate substantial economic benefits. The question is whether those purchases also contribute to domestic learning, professional development and the establishment of productive enterprises capable of serving markets beyond their national borders.
A Belizean software developer, for example, might use an internationally developed AI model to create a specialized application for agricultural management. The underlying model may be foreign, but the software architecture, customer relationships, local expertise and commercial product could belong to the Belizean enterprise. If the application becomes competitive internationally, the company may generate export revenue without having developed the foundational model itself.
The same possibility exists in tourism, logistics, healthcare administration, education and numerous other industries. The commercial opportunity is not necessarily to replicate every component of the technological infrastructure. It is to identify where domestic knowledge and specialized applications can create distinctive value.
However, ownership alone is insufficient. A locally developed product must still satisfy customer requirements, maintain technical reliability, protect relevant data and compete economically with alternatives. Governments should not confuse domestic origin with commercial viability, just as businesses should not assume that adopting foreign technology automatically represents strategic dependence.
The objective should be to develop the capacity to make informed choices rather than become permanently dependent upon whatever technology happens to be available.
The Caribbean Has Been Here Before
Our region’s economic history provides considerable reason to examine this transition carefully. Many Caribbean economies developed within international commercial systems that concentrated particular forms of production, ownership and financial influence outside the territories where economic activity occurred. Agriculture, extractive industries, shipping and tourism connected regional economies to international markets, but the distribution of value depended heavily on ownership arrangements and the structure of trade.
Independence and subsequent economic development altered many of those relationships, although important dependencies remain. Countries continue importing manufactured goods, energy products, machinery and digital technologies while relying on tourism, agricultural exports, services, remittances and other activities to generate foreign exchange. These arrangements vary substantially across the region, and it would be inaccurate to describe every Caribbean economy through a single economic model.
Nevertheless, the broader historical lesson is relevant. Participating in an industry does not necessarily mean controlling the most valuable components of its commercial structure.
A country may attract international visitors without owning the principal booking platforms through which those visitors arrange their travel. It may export agricultural products while importing the equipment and technologies required to process them. It may develop a talented workforce whose most commercially valuable opportunities remain concentrated in foreign institutions.
Artificial intelligence could reproduce some of these relationships if regional participation is limited primarily to purchasing services and supplying data, labor or commercial access.
But the technology could also provide opportunities to alter aspects of that arrangement. The declining cost of certain intellectual tasks may reduce barriers facing entrepreneurs who previously required substantial resources to obtain technical assistance, develop software or conduct preliminary research. A small business may become capable of performing activities that once required a larger administrative organization.
Those possibilities should not be romanticized. AI cannot eliminate the need for capital, physical infrastructure, competent management or access to customers. Nor does technological accessibility automatically overcome inequalities in education, connectivity and institutional support. The benefits will depend on whether people possess the resources and opportunities necessary to act upon the information these systems provide.
Still, the economic proposition deserves serious attention.
For countries that have historically encountered technological progress primarily through imported products, the possibility of developing commercially useful intellectual capabilities at lower cost represents an opportunity worth investigating.
Belize: A Small Country With a Much Larger Question
Belize provides an interesting case study because its relatively small population and domestic market create limitations that differ substantially from those confronting major industrial economies. Businesses cannot necessarily rely on large numbers of domestic customers to support expensive technological development. Public institutions must allocate limited resources among competing priorities involving infrastructure, education, healthcare, security and economic development.
Yet the country’s scale may also create opportunities for carefully targeted implementation. A government department, agricultural cooperative or specialized business may be able to identify a defined operational problem and evaluate whether an AI application can improve performance. The objective need not be an elaborate national technological transformation announced through an expensive public campaign. It may begin with practical improvements whose economic benefits can be measured.
Agriculture offers one example. Farmers must make decisions involving weather conditions, planting schedules, pests, fertilizers, equipment, financing and access to markets. AI-assisted systems could potentially help organize relevant information, identify risks and support planning. Their usefulness would depend on reliable agricultural data, local expertise and the ability of farmers to act upon the recommendations.
Tourism provides another opportunity. Belize possesses internationally recognized natural and cultural attractions, but tourism enterprises operate within a competitive global industry. AI applications could support customer communication, multilingual services, operational planning and the analysis of visitor preferences. Yet these tools cannot compensate for inadequate transportation, poor service quality or the absence of commercially attractive experiences.
Public administration may present opportunities of its own. Government institutions process applications, maintain records, administer regulations and communicate with citizens. Appropriate automation could potentially reduce administrative delays and improve access to information. But the adoption of such systems requires attention to privacy, cybersecurity, legal authority, accessibility and the accuracy of decisions affecting individuals.
These examples illustrate a broader point. Artificial intelligence becomes economically meaningful when it improves the performance of institutions and enterprises that already possess responsibilities within society.
Belize does not need to begin by asking how to become another Silicon Valley. It needs to identify where technological capabilities can strengthen the productive activities, institutions and professional opportunities upon which its economy depends.
Jamaica, Barbados and Trinidad and Tobago: Different Economies, Different Opportunities
A regional strategy must recognize that Caribbean countries possess different economic structures and technological capabilities. Jamaica’s economy includes tourism, services, agriculture, logistics and a substantial business-process outsourcing industry. Barbados possesses an established international services sector and has pursued digital modernization initiatives. Trinidad and Tobago’s industrial capabilities and energy sector create a different set of opportunities involving manufacturing, engineering and energy-intensive activities.
These distinctions matter because AI adoption will not produce identical benefits across every economy.
For Jamaica, the relationship between artificial intelligence and business-process outsourcing deserves particular attention. AI systems are increasingly capable of performing administrative and customer-service tasks traditionally associated with parts of the outsourcing industry. That creates potential pressure on particular forms of employment while also introducing opportunities to develop higher-value services involving technology management, specialized support and more sophisticated professional activities.
Barbados may examine opportunities involving financial services, digital administration, tourism and specialized professional work. The commercial benefits would depend on regulatory conditions, technical capacity and the ability to establish internationally competitive services.
Trinidad and Tobago could explore applications involving industrial maintenance, energy operations, engineering analysis and manufacturing efficiency. These opportunities differ from those confronting economies with fewer established industrial facilities, demonstrating why regional technological development should reflect existing productive capabilities rather than impose identical priorities everywhere.
There is also the question of cooperation. Smaller economies may benefit from sharing certain research resources, technical standards, training programs and specialized infrastructure. However, cooperation requires governance arrangements capable of determining how costs, responsibilities and benefits are distributed.
Regional unity is a valuable principle, but it does not eliminate the practical requirements of financing and institutional accountability.
Latin America Is Not Starting From Zero
The conversation becomes considerably more complex when we expand beyond the Caribbean into Latin America. Brazil, Mexico, Colombia, Chile, Argentina and other countries possess substantial universities, research communities, technology enterprises and industrial capabilities. Their opportunities and constraints differ significantly from those of smaller Caribbean states.
Brazil’s agricultural, industrial and financial sectors create opportunities for AI applications at considerable scale. Mexico’s integration into North American manufacturing supply chains introduces questions concerning industrial automation, logistics and technological competitiveness. Colombia’s digital services and entrepreneurial sectors provide another set of possibilities, while Chile’s mining industry and energy infrastructure create opportunities involving resource management and industrial efficiency.
These countries are not merely waiting for technology to arrive. Across Latin America, researchers, businesses and public institutions have already been developing AI-related capabilities and policy frameworks. The challenge is understanding whether those activities can produce broader improvements in productivity, domestic innovation and economic opportunity.
There is also a risk of concentrating attention exclusively on the region’s largest economies. Smaller countries may develop competitive capabilities in specialized industries without replicating the technological infrastructure of Brazil or Mexico. A country with expertise in a particular agricultural product, environmental system or professional service may identify applications whose commercial value extends internationally.
The relevant distinction is between scale and specialization.
Large economies possess advantages in domestic demand, research financing and industrial infrastructure. Smaller economies may possess opportunities to develop focused products, specialized knowledge and commercially valuable services that do not require enormous domestic markets.
The challenge is identifying those opportunities through evidence rather than assuming that every country can become competitive in every technological field.
What China’s Social Anxiety Should Teach Us
Yi-Ling Liu’s discussion of Chinese society introduces another dimension that deserves attention beyond technological competition. Her examination of economic stratification, intense educational competition and the lying flat movement illustrates how technological advancement can coexist with dissatisfaction concerning work, opportunity and social mobility.
The phenomenon known as tang ping, or lying flat, emerged as an expression of resistance to intense social and economic expectations. It reflects, among other things, frustration with the pressure to compete continuously for employment, housing and conventional measures of success. Its significance extends beyond the particular circumstances of China because similar anxieties exist in societies confronting expensive living conditions and uncertain economic mobility.
Latin America and the Caribbean possess their own versions of these pressures. Young people may pursue education without finding employment corresponding to their qualifications. Others migrate because opportunities available abroad appear more attractive than those within their home countries. Entrepreneurs encounter difficulties obtaining financing, while households face pressures involving housing, transportation and basic living expenses.
Artificial intelligence will enter economies already shaped by these conditions.
If businesses use AI primarily to reduce labor expenses without developing new productive opportunities, some workers may experience greater insecurity. If technological adoption improves productivity and supports the development of competitive industries, new opportunities may emerge. The distribution of these outcomes will depend on the industries involved, labor-market institutions, education and the policies governing technological transition.
This is why the economic discussion cannot be reduced to the number of AI systems a country adopts.
A society may become technologically sophisticated without becoming substantially more prosperous for the majority of its population. Conversely, carefully selected technological applications may improve services and productivity without requiring dramatic changes to every aspect of economic life.
The relevant question is whether technological development contributes to a society in which people possess greater economic capacity and meaningful opportunities.
Artificial Intelligence Cannot Replace an Economy
One of the more dangerous assumptions surrounding AI is that technological capability can compensate for almost every structural weakness affecting economic development. A government may acquire sophisticated analytical systems while continuing to struggle with unreliable electricity, inadequate transportation and administrative inefficiency. A business may generate impressive strategic plans through AI while lacking the financing necessary to implement them.
The distinction between knowledge and execution remains fundamental.
An AI system may help design an irrigation network, but farmers still require the equipment, water infrastructure and financing necessary to construct it. A model may identify opportunities to improve port efficiency, but implementation may require physical upgrades, regulatory changes and cooperation among multiple institutions. A technology may assist medical professionals with administrative work, but healthcare systems still require trained personnel, equipment and functioning facilities.
The danger is confusing access to intelligence with possession of productive capacity.
For developing economies, that distinction should influence how technological investments are evaluated. Governments and businesses should consider the complete cost of implementation, including infrastructure, training, maintenance, cybersecurity and the organizational changes necessary to achieve useful results.
A successful AI project should not be measured exclusively by whether the technology functions. It should be evaluated according to whether the intended institution or enterprise performs better because the technology was introduced.
That may involve lower operating costs, improved service quality, increased output or greater resilience against disruptions. The relevant indicators will differ according to the application, but they should be established before substantial resources are committed.
Technological sophistication is not a substitute for economic discipline.
The Question of Ownership
As a founder developing enterprises across design, technology and other industries, I have become increasingly interested in the relationship between intellectual property and institutional value. A business may generate considerable economic activity without possessing the assets through which the most valuable components of that activity are controlled.
Artificial intelligence makes this distinction particularly important.
An enterprise may depend on an external model provider while owning the software, workflows, customer relationships and specialized data associated with its own product. Another may develop proprietary models but remain dependent on external computing infrastructure. These arrangements introduce different forms of commercial exposure and control.
There is no universal requirement that every company own every technological component it uses. Commercial partnerships, licensing and external infrastructure can be economically efficient. The important consideration is understanding which dependencies are acceptable and which may compromise the enterprise’s ability to operate, compete or preserve its intellectual property.
For Latin American and Caribbean businesses, the issue extends beyond technical architecture. It concerns whether domestic entrepreneurs can establish companies whose products generate revenue internationally, attract investment and create professional opportunities within their economies.
It also concerns the legal and institutional environment supporting those businesses.
Here is the remaining portion of the HYDESMAN POST article, continuing directly from the unfinished sentence in “The Question of Ownership.” It completes the discussion of intellectual property, examines regional cooperation, governance and technological dependence, and concludes with a strategic outlook for 2027–2031.
Intellectual property protection, access to financing, reliable commercial contracts, technical education and predictable regulation influence whether innovative enterprises can develop into sustainable institutions. A talented software engineer may create a commercially valuable application, but transforming that application into an international business requires capabilities extending beyond programming. The entrepreneur must establish ownership rights, protect customer information, finance development, negotiate contracts and build relationships with markets capable of supporting continued growth.
This is where the discussion of artificial intelligence becomes a discussion about economic institutions. If Latin America and the Caribbean want to participate meaningfully in the emerging technological economy, they must consider the environments in which domestic businesses are established, financed and allowed to compete. Access to sophisticated AI systems may lower certain barriers to innovation, but it does not eliminate the commercial and institutional requirements necessary to build valuable enterprises.
The question is therefore not whether every Caribbean country should develop its own version of the world’s largest artificial intelligence companies. It is whether a developer in Belize, Jamaica, Barbados or Trinidad and Tobago can build a competitive technological enterprise without being structurally disadvantaged by the absence of appropriate financing, infrastructure and institutional support. Those conditions will influence whether the region develops lasting technological capabilities or remains primarily a market for products developed elsewhere.
CARICOM and the Possibility of Regional Technological Capacity
The Caribbean’s relatively small national markets introduce an important question concerning regional cooperation. Individual countries may struggle to finance specialized computing infrastructure, research facilities or extensive technical training programs. Collectively, however, participating states may possess greater opportunities to coordinate investment, establish shared standards and develop capabilities that would be difficult to sustain independently.
CARICOM provides an existing institutional framework through which certain forms of cooperation could be explored. Regional universities, development institutions, public agencies and private enterprises could contribute specialized expertise, financing and operational capacity. The objective would not necessarily be to create a single centralized artificial intelligence authority, but to identify functions where cooperation produces measurable advantages over fragmented national initiatives.
Consider the development of shared technical resources. Universities and researchers across the region may benefit from access to specialized computing infrastructure, common research datasets and collaborative training programs. Public institutions could potentially cooperate on cybersecurity standards, procurement guidance and the responsible use of AI in administrative services. Businesses might benefit from regional professional networks and mechanisms connecting technological expertise with commercial opportunities.
However, regional cooperation should not be confused with the assumption that every participating country possesses identical interests. Larger economies may have different priorities from smaller island states, while countries with established industrial capabilities may require technologies that are less relevant to tourism-dependent economies. Financing arrangements would need to account for differences in resources, anticipated benefits and institutional responsibilities.
There is also the question of governance. Who would own shared infrastructure? Which institutions would determine access? How would commercially sensitive information be protected, and what mechanisms would prevent resources from becoming concentrated among a small number of favored participants? These questions are not obstacles to cooperation, but they must be resolved if regional initiatives are to produce durable economic value.
The Caribbean has considerable experience establishing regional institutions and negotiating common objectives. Artificial intelligence introduces another opportunity to examine where collective action may strengthen national capabilities. The appropriate measure of success would be whether participating countries acquire useful resources, expertise and productive opportunities that justify the investment.
The Development Banks Have a Role Beyond Financing Technology Purchases
International and regional development institutions already influence infrastructure, education, public administration and private-sector investment across Latin America and the Caribbean. The Caribbean Development Bank, Inter-American Development Bank and other financing institutions possess experience supporting projects intended to improve economic capacity. Artificial intelligence introduces questions about how those existing responsibilities should evolve.
One possibility is that development financing increasingly supports the acquisition of AI-enabled systems for public institutions and businesses. Such investments may be useful where they address clearly identified operational requirements. However, financing technology purchases without ensuring institutional readiness can produce expensive systems that fail to deliver anticipated improvements.
A more comprehensive approach would examine the complementary investments required for successful implementation. Reliable electricity, broadband connectivity, technical education, cybersecurity, data management and institutional reform may be as important as the AI applications themselves. Development institutions could help governments evaluate these requirements before committing substantial resources.
There is also an opportunity to support domestic technological enterprises. Entrepreneurs developing commercially useful applications may require access to early-stage financing, professional services and international markets. Traditional lending arrangements may not adequately accommodate businesses whose principal assets consist of software, intellectual property and specialized knowledge rather than conventional physical collateral.
That does not mean development banks should finance speculative technology ventures without appropriate commercial scrutiny. Public and development resources require disciplined assessment of risks, expected benefits and the capacity of recipients to execute their plans. The objective should be to address identifiable financing gaps while maintaining standards appropriate to the use of institutional capital.
For smaller economies, technical assistance may sometimes be more valuable than a large financing commitment. Helping a government establish reliable procurement procedures or assisting businesses with intellectual property and commercial planning can strengthen the environment in which future investments occur. The development of technological capability is an institutional process, not simply a matter of purchasing equipment.
Who Controls the Data?
The question of data deserves particular attention because artificial intelligence systems depend upon information for training, evaluation and operation. Governments maintain records concerning citizens and public services, while businesses collect information involving customers, transactions and commercial activities. Researchers may work with agricultural, environmental, linguistic and scientific datasets whose value increases when they can be organized and analyzed effectively.
These resources introduce opportunities for improving services and developing specialized technologies. They also create responsibilities concerning privacy, security, consent and the lawful use of information. The fact that a dataset could improve an AI application does not automatically establish that an institution possesses the authority to use it for that purpose.
For Caribbean and Latin American countries, the issue becomes especially important when public institutions contract with international technology providers. Governments should understand where information will be processed, which parties may access it, how long it will be retained and whether contractual arrangements permit secondary uses. They should also consider how services would continue if a provider changes its commercial terms or discontinues a product.
Data sovereignty is sometimes discussed as though every country must physically store all information within its borders. In practice, the appropriate arrangements depend on the sensitivity of the information, applicable law, technical security and the operational requirements of the service. Domestic storage does not automatically guarantee security, just as international infrastructure does not necessarily eliminate meaningful national control.
The more important objective is to establish enforceable rights, reliable safeguards and institutional competence. Governments must be capable of understanding the systems they procure and ensuring that their use remains consistent with legal obligations and the interests of the people affected.
There is also a cultural dimension. Latin America and the Caribbean possess linguistic traditions, historical records and forms of local knowledge that may be poorly represented within internationally developed AI systems. Indigenous languages, Caribbean Creoles, Garifuna and other linguistic communities raise questions concerning preservation, representation and access to technological tools capable of supporting their continued use.
Developing appropriate language resources could contribute to education, cultural preservation and the creation of specialized applications. But such initiatives require participation from the communities whose languages and knowledge are involved. Cultural information should not simply be collected as another commercial resource without considering consent, attribution and the interests of its custodians.
Artificial intelligence may expand access to knowledge, but the institutions developing and deploying it must remain accountable for how that knowledge is obtained and used.
The Human Economy Must Remain Part of the Equation
One of the most consequential questions raised by Yi-Ling Liu’s discussion concerns the relationship between technological progress and human participation in economic life. Societies do not exist simply to maximize output. They also provide the institutions through which people earn livelihoods, develop skills, establish relationships and participate in their communities.
Artificial intelligence may alter the organization of work across numerous industries. Certain administrative tasks may become less expensive to perform, while new responsibilities emerge involving the supervision, maintenance and application of technological systems. The resulting effects will vary according to occupations, industries and the pace of adoption.
For Latin America and the Caribbean, the implications are particularly important because many economies already confront challenges involving employment quality, informal work, migration and access to professional opportunities. Technological adoption could improve productivity while also creating pressure on workers whose tasks become increasingly automated. The appropriate response requires examining both possibilities rather than assuming that innovation will automatically produce either widespread unemployment or universal prosperity.
Education systems will have an important role, but the challenge extends beyond teaching students to use AI applications. Workers may require stronger foundations in mathematics, communication, technical reasoning, professional judgment and the ability to evaluate information critically. Businesses will need people capable of identifying appropriate applications, managing technological systems and understanding the limitations of automated processes.
There is also a responsibility to preserve meaningful human expertise. An institution that relies entirely on automated systems may gradually lose the professional knowledge necessary to evaluate their performance or respond when they fail. A country that imports increasingly sophisticated technology without developing domestic expertise may become more efficient in certain activities while remaining dependent on external providers for critical decisions.
This introduces a distinction between automation and capability development. Automation reduces or reorganizes the human effort required to perform particular tasks. Capability development strengthens the ability of people and institutions to understand problems, make decisions and execute useful work. The two can complement one another, but they are not interchangeable.
The objective should not be to preserve every existing task regardless of technological progress. Economic development has always involved changes in the organization of production and employment. The more consequential question is whether societies possess mechanisms through which people can acquire new capabilities and participate meaningfully in the opportunities created by those changes.
Artificial intelligence should expand what people and institutions can accomplish. Its economic success should not be measured exclusively by how many human responsibilities can be eliminated.
The Geopolitical Question: Must We Choose a Side?
The competition between the United States and China introduces another complication for countries attempting to develop technological capabilities. Advanced computing infrastructure, semiconductor supply chains, cloud services and AI platforms increasingly intersect with national security and international trade policy. Governments may encounter pressure to consider the geopolitical implications of their technology partnerships alongside ordinary commercial requirements.
Latin America and the Caribbean maintain important relationships with multiple international partners. The United States remains a significant commercial and financial partner for many countries, while China participates in regional trade, infrastructure investment and technology markets. European institutions and companies also contribute to the technological and regulatory environment.
The existence of these relationships does not require every country to adopt an identical diplomatic position. National governments must evaluate technology arrangements according to their economic requirements, legal obligations, security considerations and the terms offered by potential partners.
However, technological dependence can introduce vulnerabilities. A country relying heavily on a single provider may face difficulties if access becomes restricted, prices increase or international regulations change. Diversification may reduce certain exposures, although maintaining multiple systems can introduce additional costs and operational complexity.
The appropriate approach is therefore not simply to purchase technology from whichever international partner offers the most attractive immediate price. Governments and businesses should consider reliability, security, interoperability, contractual protections and the long-term costs of maintaining their systems.
There may also be opportunities for scientific and commercial cooperation that extend beyond geopolitical competition. Researchers, universities and businesses can exchange knowledge and develop useful applications through relationships that do not necessarily require alignment with every political objective of their respective governments.
Liu’s discussion of continued intellectual exchange between American and Chinese researchers provides a reminder that technological development is not exclusively conducted through state rivalry. Scientific collaboration, professional relationships and commercial exchange remain important components of innovation.
For Latin America and the Caribbean, international partnerships should be evaluated according to what they enable the region to accomplish. The objective is not technological isolation, but the capacity to participate in international markets without surrendering the ability to make informed decisions about critical systems.
Strategic Intelligence Scaffolding: What Should We Actually Measure?
The discussion becomes more useful when translated into measurable economic and institutional questions. Through the principles of Strategic Intelligence Scaffolding, the objective is to examine artificial intelligence not as an isolated technology, but as one component of a broader system involving capital, infrastructure, education, governance, trade and productive activity.
A country may announce an ambitious AI strategy without possessing the resources necessary to implement it. Another may make relatively modest investments that produce measurable improvements in particular industries. The distinction between these outcomes cannot be established through announcements alone.
Several indicators deserve attention over the coming years. These include the reliability and cost of electricity, access to broadband connectivity, the availability of technical expertise, investment in domestic technology enterprises, the quality of public-sector data and the measurable productivity effects of AI adoption.
The structure of commercial ownership also matters. Governments and development institutions should examine whether local enterprises are creating intellectual property, generating export revenue and developing professional opportunities. The number of AI subscriptions purchased within an economy is not necessarily a meaningful measure of its technological development.
Institutional performance deserves similar scrutiny. A government department implementing AI should be capable of demonstrating whether the technology has reduced processing times, improved accuracy or increased access to services. Where systems fail to produce measurable benefits, institutions should be willing to revise or discontinue them rather than preserve investments simply because they have already been made.
There is also the question of distribution. Productivity improvements may benefit businesses, workers, consumers and governments in different ways. An increase in output does not automatically establish that household purchasing power or employment conditions have improved. Evaluating the economic consequences of AI therefore requires attention to who receives the benefits and who bears the costs.
These indicators cannot eliminate uncertainty, but they can improve the quality of decisions. They allow technological development to be assessed according to economic performance rather than enthusiasm, geopolitical rhetoric or the prestige associated with adopting advanced systems.
The Next Five Years: Three Possible Futures for the Region
The period from 2027 through 2031 provides a useful horizon for examining how artificial intelligence might influence Latin America and the Caribbean. Five years is sufficiently long for investments, education initiatives and institutional reforms to begin affecting economic performance, while remaining close enough to present conditions to support meaningful planning.
Nevertheless, technological development and international economic conditions remain uncertain. The appropriate approach is to examine conditional scenarios rather than assume that a particular outcome is inevitable.
Scenario One: The Region Becomes a Larger Market for Imported Intelligence
Under this scenario, AI adoption expands across businesses and public institutions, but domestic technological capability develops more slowly. Governments purchase administrative systems, enterprises subscribe to foreign platforms and consumers increasingly rely on internationally developed services.
These technologies may produce genuine improvements in productivity and convenience. However, domestic participation in software development, intellectual property ownership and specialized technological services remains relatively limited. Much of the commercial value associated with the underlying systems continues to accrue to international providers.
The region becomes more technologically connected without necessarily developing a proportionate increase in domestic technological production.
This scenario would not represent complete economic failure. Imported technology can generate substantial benefits, particularly when it improves the performance of existing industries. The limitation would be the extent to which regional economies remain dependent upon external providers for critical capabilities and opportunities for commercial expansion.
Scenario Two: A Regional Economy of Applied Intelligence Emerges
Under a second scenario, governments, businesses and educational institutions coordinate selected investments in infrastructure, technical training and commercially useful applications. Domestic enterprises develop specialized products for agriculture, tourism, logistics, financial services and public administration.
Regional cooperation supports particular research and professional capabilities, while international partnerships provide access to technologies that would be inefficient to reproduce independently. Development institutions help address financing and institutional constraints, and businesses increasingly compete through specialized expertise rather than attempting to replicate the largest global technology companies.
The resulting economy would still depend upon international technology providers in numerous areas. However, a greater proportion of the value associated with implementation, customization, intellectual property and specialized services could be generated within the region.
This outcome would depend on sustained investment, institutional competence and the ability of businesses to develop commercially viable products. It would not emerge automatically from the availability of AI systems.
Scenario Three: Technological Adoption Deepens Existing Inequalities
A third possibility involves uneven adoption across countries, industries and social groups. Better-capitalized businesses acquire sophisticated systems, while smaller enterprises struggle with financing, technical expertise and infrastructure limitations.
Governments with stronger administrative capabilities improve certain services, while institutions facing structural weaknesses experience limited benefits from technology purchases. Workers possessing relevant skills may encounter new opportunities, while others face pressure from automation without adequate access to retraining or alternative employment.
Under these conditions, technological progress could coexist with widening differences in productivity, income and institutional performance.
This scenario is not mutually exclusive with the others. Some countries may develop competitive technological enterprises while others remain primarily consumers of imported systems. Even within individual economies, successful applications may emerge alongside industries experiencing significant disruption.
The purpose of examining these possibilities is not to assign a predetermined future to the region. It is to identify the conditions that could influence how technological development translates into economic outcomes.
What Could Change the Trajectory?
Several developments could materially influence the region’s technological future.
Improved access to reliable electricity and telecommunications infrastructure could reduce constraints affecting businesses and public institutions. Advances in the efficiency of AI models may lower the computing resources required for particular applications, potentially expanding opportunities for smaller enterprises. Stronger technical education and professional training could increase the number of people capable of developing and managing useful systems.
Financial innovation may also influence the pace of development. Investment mechanisms better suited to software enterprises and intellectual property could support businesses whose commercial potential is not adequately reflected through conventional collateral requirements. However, financing must remain connected to credible business models and realistic expectations concerning demand.
Institutional reform may be equally consequential. Governments that improve procurement, strengthen data governance and establish clear accountability for technology projects may obtain greater value from existing resources. Regional cooperation could reduce duplication where shared infrastructure or expertise produces measurable advantages.
International conditions will also matter. Changes in semiconductor availability, cloud-service pricing, trade restrictions and geopolitical relationships may influence the cost and accessibility of advanced technologies. Countries possessing diversified commercial relationships and adaptable technical systems may have more options when these conditions change.
The most consequential development may be the emergence of commercially successful regional enterprises capable of serving international markets. A small number of competitive businesses can demonstrate viable models, develop professional expertise and create opportunities for subsequent entrepreneurs. Their success would not automatically transform entire national economies, but it could contribute to the development of a more substantial technological ecosystem.
We Should Not Wait for the Future to Be Delivered
There is something revealing about the language used to describe technological change in developing economies. We frequently speak about preparing for the future, adapting to global developments and ensuring that our countries are not left behind. These are reasonable objectives, but they can also encourage a passive understanding of economic development, as though the future is something created elsewhere and eventually delivered to us.
Artificial intelligence presents an opportunity to reconsider that assumption.
Latin America and the Caribbean do not need to reproduce every technological capability developed in the United States or China. Nor should our governments pursue expensive technological projects simply to demonstrate participation in an international competition. The more important responsibility is to understand how emerging capabilities can strengthen the industries, institutions and professional opportunities upon which our economies depend.
That requires decisions concerning education, infrastructure, investment, ownership and governance. It requires businesses willing to develop commercially useful products and governments capable of evaluating technology according to measurable public benefits. It also requires recognizing that technological development is inseparable from the wider economic conditions affecting people’s lives.
Yi-Ling Liu’s discussion of American and Chinese approaches to artificial intelligence illustrates that even the world’s largest technological powers do not understand the future through identical priorities. Their strategies reflect different economic structures, political institutions, cultural expectations and development objectives.
There is no reason Latin America and the Caribbean should assume that our priorities must be identical to either.
Our countries possess different resources, industries, demographic conditions and institutional capabilities. Those differences should inform the technological strategies we pursue rather than be treated as disadvantages that must somehow be overcome through imitation. A country whose economic strengths lie in agriculture may identify opportunities different from those of an economy specializing in financial services, manufacturing or tourism. The objective should be to develop capabilities appropriate to those circumstances while remaining sufficiently adaptable to participate in an evolving international economy.
There is also a question of ambition. For generations, people across Latin America and the Caribbean have demonstrated extraordinary ingenuity in navigating limited resources, international economic pressures and institutions that have not always provided equal opportunities for advancement. That ingenuity deserves recognition, but resourcefulness should not become a permanent substitute for the infrastructure, financing and institutional support required to build competitive enterprises. We should not celebrate our ability to improvise so enthusiastically that we neglect the responsibility to create conditions in which improvisation is no longer necessary for ordinary economic survival.
As a founder developing businesses across fashion, technology and other industries, I recognize the importance of understanding how knowledge acquired in one field can become useful in another. Artificial intelligence has the potential to expand that capacity considerably, allowing individuals and organizations to approach unfamiliar problems with analytical resources that were previously difficult or expensive to obtain. But possessing access to intelligence is only the beginning. The greater challenge is translating that intelligence into functioning enterprises, productive assets, professional opportunities and institutions capable of delivering measurable value.
This is where I believe the conversation must ultimately return to people. The purpose of economic development is not simply to produce impressive technological statistics or announce increasingly sophisticated systems. It is to expand the capacity of societies to provide security, opportunity, education, healthcare, productive employment and the conditions through which individuals can pursue meaningful lives. Artificial intelligence may contribute to those objectives, but it cannot independently determine what prosperity should mean for the communities expected to live with its consequences.
The United States and China will continue making decisions according to their own economic interests, strategic priorities and institutional arrangements. Their companies will develop technologies, establish commercial relationships and compete for influence across international markets. Latin America and the Caribbean will inevitably participate in those developments, whether through investment, trade, technological adoption or the changing structure of employment.
But participation should not be confused with passivity.
We have universities, entrepreneurs, researchers, engineers, cultural institutions, agricultural producers, manufacturers and generations of people who understand the circumstances of their own societies. We possess knowledge that cannot always be reproduced through an imported technological system, particularly where economic activity depends on local environments, languages, relationships and institutional realities. The challenge is connecting those capabilities with the resources and organizational structures necessary to transform them into productive economic value.
Perhaps the most consequential question is not whether Latin America and the Caribbean can compete directly with the world’s largest artificial intelligence laboratories. It is whether we can establish economies in which our own people possess the knowledge, resources and institutional support necessary to create businesses, develop technologies and participate more fully in the wealth generated by global innovation.
That is a question of ownership, certainly, but it is also a question of education, governance, infrastructure and collective economic ambition. It requires recognizing that a country can benefit from internationally developed technology while still developing valuable capabilities of its own. It requires understanding that cooperation with larger economies need not eliminate national agency, and that technological modernization should strengthen rather than diminish the human institutions upon which societies depend.
The artificial intelligence revolution will not wait for every country to establish a comprehensive strategy. Businesses are already adopting these systems, workers are encountering their consequences, and governments are making procurement and regulatory decisions that may influence their economies for years to come. The relevant choices are therefore not confined to some distant technological future. They are being made now, through investments, contracts, educational priorities and the opportunities available to entrepreneurs.
For Latin America and the Caribbean, the responsibility is to approach those choices with greater intentionality than has sometimes characterized earlier periods of technological adoption. We should examine what the technology can accomplish, what it costs, who controls it, which institutions are prepared to use it and how its benefits can contribute to broader economic development.
We should also be willing to ask whether our existing economic arrangements are capable of converting new technological opportunities into prosperity. If they are not, acquiring more sophisticated systems will not necessarily resolve the underlying problem. Technology can expand the possibilities available to an economy, but the institutions through which those possibilities are realized remain essential.
The world may continue describing artificial intelligence as a race between two superpowers. That framing captures an important dimension of technological competition, but it does not adequately describe the interests of the billions of people whose futures will also be influenced by the outcome.
Latin America and the Caribbean do not need to win someone else’s race to develop a meaningful technological future. We need to understand our own economic objectives, develop the capabilities required to pursue them and establish institutions capable of preserving the value we create.
Comments