Volume 2: AI as a Strategic Industry

The internet always seemed weightless, a realm of pure information floating above the physical world of nations, borders, and power. For decades, we treated digital infrastructure as if it existed nowhere and everywhere at once. But as artificial intelligence scales from research curiosity to industrial foundation, that illusion is shattering. AI has mass. It requires land, energy, minerals, and factories. And like every strategic industry before it, from oil to semiconductors, whoever controls AI's physical foundations will shape the global order for generations.

Artificial intelligence is now a strategic industry, joining the exclusive ranks of sectors that determine which nations rise and which fall behind. It rivals finance, healthcare, and heavy industry in projected GDP impact, but its significance runs deeper. Strategic industries aren't just economically important; they're the ones governments cannot afford to lose, the ones that create dependencies others can exploit, and the ones that multiply power across every other sector of society.

For the Americas, this transformation arrives at a pivotal moment. The United States seeks to maintain technological dominance while reshoring critical manufacturing. Canada navigates between research excellence and economic capture. Mexico faces a choice between competing spheres of influence. And across the rest of Latin America, nations must decide what role they wish to play as active participants in the AI economy.

From Technology Sector to Strategic Industry

What separates a strategic industry from merely an important one? Three characteristics define the difference: the ability to multiply economic power across all other sectors, direct implications for national security and sovereignty, and the creation of dependencies that can be weaponized in times of conflict.

Oil demonstrated this pattern throughout the 20th century. Nations that controlled production and refining shaped global politics. Those dependent on imports found their foreign policy constrained by energy security concerns. The 1973 oil embargo showed how quickly a strategic commodity could become a geopolitical weapon. Semiconductors followed a similar trajectory, what began as components in consumer electronics became the foundation of everything from smartphones to missile systems, leading to today's battles over chip manufacturing capacity and export controls.

Artificial intelligence is following this same arc, but faster and with potentially greater consequences. Unlike oil, which powers specific sectors, or semiconductors, which enable digital technology, AI is fundamentally horizontal. It reaches into agriculture and logistics, healthcare and defense, financial services and manufacturing. A nation that falls behind in AI capabilities doesn't just lose competitiveness in one industry; it risks falling behind across the entire economy.

This is why the AI industry is pulling away from what we traditionally thought of as the technology sector and establishing itself as something distinct. Software companies, cloud providers, and semiconductor manufacturers are converging into an integrated industrial base that governments increasingly view as too important to leave entirely to market forces. The scale of investment required, in computing infrastructure, energy capacity, specialized facilities, and human capital, demands coordination between public and private sectors that resembles wartime mobilization more than typical industrial policy.

The New Cold War of Compute

If AI's strategic importance were unclear, the intensity of great power competition would resolve any doubt. The United States and China have embarked on what can only be described as a new Cold War of Compute, a contest to dominate the foundational technologies of artificial intelligence with implications as profound as the nuclear arms race or the space race of the previous century.

The architecture of this competition reveals much about how both powers understand AI's strategic value. The United States has constructed the Chip4 alliance with Japan, South Korea, and Taiwan, creating an integrated semiconductor supply chain explicitly designed to exclude China from access to the most advanced chips. These aren't chips for consumer electronics; they're the specialized processors required for training and deploying cutting edge AI systems. By controlling this chokepoint, the United States aims to maintain a generational advantage in AI capabilities.

China has responded with its own strategy of building parallel infrastructure through the Digital Silk Road, an initiative that combines direct investment in developing nations' digital infrastructure with the export of Chinese technology standards and platforms. Through the Digital Silk Road, China offers countries what the United States and its allies often cannot or will not: patient capital, minimal political conditions, and comprehensive technology packages that don't require navigating multiple vendors and complex integration.

For the Americas, this competition transforms the hemisphere from a region that might have remained peripheral to AI geopolitics into a central theater. Brazil, Ecuador, and Peru have all engaged with the Digital Silk Road, accepting Chinese investment in telecommunications infrastructure, data centers, and surveillance technology. These aren't mere commercial relationships; they're strategic positioning that could shape everything from data flows to standards adoption to security vulnerabilities for decades.

The stakes became visceral when advanced semiconductors that had been specifically designed to comply with 2022 U.S. export control guidance were abruptly banned from export to China in a 2025 policy reversal. Companies and nations that had planned around one set of rules suddenly faced another. The message was unmistakable: in the competition for AI dominance, the rules can change as quickly as administrations turn over, and what seemed like secure commercial relationships can evaporate overnight for national security reasons.

Strategic Positioning Across the Americas

How nations in the Western Hemisphere respond to this new strategic landscape will largely determine their economic trajectories and geopolitical relevance for the remainder of the century.

The United States has made its intentions explicit through a sequence of major policy initiatives. The CHIPS Act of 2022 committed over $50 billion to rebuilding domestic semiconductor manufacturing capacity. The Stargate Project launched in 2025 represents the most ambitious AI infrastructure initiative in history. The AI Action Plan released in July 2025 lays out a comprehensive vision: the United States will not merely lead in AI software and models but will build the complete industrial base, from chip fabrication to data center construction to power generation necessary for AI dominance.

This represents a dramatic shift from the globalized supply chain model that dominated tech industry thinking for three decades. Both the Biden and Trump administrations, despite deep disagreements on nearly everything else, agreed on this fundamental point: AI is too strategically important to depend on supply chains that route through geopolitical rivals or even uncertain allies. The goal is nothing less than end-to-end domestic capacity for the full AI stack, supplemented by tightly integrated partnerships with trusted allies. A small scale problem is arising in the court of public opinion within the United States. This infrastructure buildout is having effects on communities adjacent to it. The digital colonial path may be grim, we would see some of the polluting facilities relocated to nations desperate for foreign investment in the developing world.

Canada is pursuing what might be called a hybrid strategy, trying to capture the best elements of both the American and European approaches. Canada wants the economic growth and security partnership that comes with integration into U.S.-led AI development, but it also values the regulatory frameworks and ethical guardrails that Europe has pioneered. The challenge is whether this middle path is sustainable or whether the intensity of great power competition will eventually force more definitive alignment.

Canada's greatest offer to the artificial intelligence cause is world-class AI research capabilities but it also represents its greatest strategic vulnerability. Canadian universities and research institutions have produced foundational breakthroughs in deep learning and other AI techniques, but capital, infrastructure, and markets are concentrated south of it’s border. The risk is that Canada will continue to generate innovation that primarily benefits American companies and Chinese researchers, capturing neither the economic returns nor the strategic advantages that should flow from its intellectual contributions.

Mexico finds itself in an enviable but complicated position. Proximity to the United States and deep manufacturing integration offer enormous opportunities as American companies seek to nearshore critical supply chains. But Mexico also maintains significant economic ties with China, particularly in telecommunications infrastructure where Chinese vendors have substantial market share. The United States has made clear that participation in its AI industrial base will require partners to decouple from Chinese influence in sensitive sectors.

Mexico's response reveals sophisticated strategic thinking. The country is moving forward with Cotilicue, a supercomputer that would be Latin America's largest and most capable. This isn't merely about computational capacity; it's a statement of intent to develop indigenous AI capabilities rather than becoming entirely dependent on either the United States or China. Mexico is positioning itself as a genuine partner in North American AI development, not just a manufacturing platform or a passive consumer of others' technology.

Latin America more broadly faces perhaps the most difficult strategic calculus. Nations across Central and South America largely lack the resources for full-stack AI development. They don't have domestic semiconductor industries, their energy infrastructure often struggles with current demand, and their technology sectors,where they exist, are small relative to U.S. or Chinese giants.

Yet these nations possess assets that make them relevant to the AI competition: critical mineral deposits essential for chip manufacturing and data center construction, geographic positions that matter for undersea cables and satellite communications, and markets large enough to give them leverage in choosing between competing technology ecosystems. The question is whether they can convert these assets into genuine influence or whether they'll find themselves once again in a familiar pattern, exporting raw materials and cheap labor while others capture the value-added portions of the supply chain.

The Industrial Policy Toolkit

Understanding how nations are competing for AI supremacy requires examining the specific mechanisms they're deploying. Gone are the days when technology development could be left primarily to market forces and venture capital. AI has joined the ranks of industries where government policy, subsidy, and strategic coordination shape outcomes as much as innovation and entrepreneurship.

Subsidies have become the most visible tool. The United States isn't merely offering tax breaks; it's committing to direct funding for chip fabrication plants, data center construction, and AI research facilities. These subsidies aren't distributed to whoever applies; they're strategically targeted to build specific capabilities that policymakers have identified as critical. The goal is to accelerate development that market forces alone would accomplish too slowly or not at all, if security concerns or international competition make waiting unacceptable.

Export controls represent the flip side of the same strategic coin. If subsidies accelerate development of domestic capabilities, export controls prevent rivals from accessing the same technologies. The semiconductor export restrictions targeting China are the most prominent example, but they're part of a broader pattern. Advanced AI chips, manufacturing equipment for producing those chips, and even the software tools used to design them are all subject to controls that didn't exist five years ago.

These controls carry significant costs. They disrupt established commercial relationships, invite retaliation, and create uncertainty that makes long-term planning difficult for companies caught in the middle. The 2025 reversal on previously approved chip exports illustrates the problem: what businesses thought was settled policy proved negotiable when strategic calculations shifted. This unpredictability itself becomes a weapon in great power competition, as companies and nations must constantly assess whether their supply chains might be disrupted by the next policy change.

Industrial policy encompasses the broader coordination of research priorities, infrastructure investment, workforce development, and regulatory frameworks. Data centers, once viewed as mundane backend infrastructure, have become objects of national strategy. Countries are competing to attract these facilities with incentives that would have seemed excessive for heavy manufacturing in an earlier era. But data centers are where AI models train and deploy; they're the factories of the information age, and nations that host them gain not just economic activity but potentially strategic advantage in an AI-dependent world.

Perhaps no statistic better captures the vulnerability driving this industrial policy frenzy than this: over 70% of critical minerals essential for modern technology from smartphones to fighter jets to AI servers are either mined or refined in China. For any nation seeking to avoid dependence on a geopolitical rival, this concentration represents an unacceptable risk. It explains why mineral supply chains have joined semiconductors and AI software as objects of strategic concern and reshoring efforts.

Strategic Choices for the Hemisphere

The question facing nations across the Americas isn't whether to engage with AI, that ship has sailed. The question is how to position themselves within an industry being actively shaped by great power competition.

For smaller Latin American nations, the temptation of the best-of-both-worlds strategy is obvious: maintain good relations with the United States for security and market access while preserving economic ties with China for investment and trade. Europe has pursued exactly this approach, trying to remain friendly with both sides of the U.S.-China divide while developing its own regulatory framework as a form of strategic autonomy.

But the intensity of the New Cold War of Compute may make this balancing act increasingly difficult. Both the United States and China are demanding more definitive alignment, particularly in sensitive sectors like telecommunications, cloud computing, and AI infrastructure. A nation might find that accepting Chinese investment in 5G networks closes doors to participation in U.S.-led AI initiatives, or that integrating with American technology platforms requires severing relationships with Chinese vendors.

The risk isn't merely geopolitical pressure; it's economic in nature. AI development benefits enormously from network effects and ecosystem integration. A nation that tries to use both Chinese and American AI infrastructure may find itself struggling with compatibility issues, unable to fully participate in either ecosystem. The choice of which AI platforms to build around, which cloud providers to depend on, and which technology standards to adopt will shape economic trajectories for decades.

Yet there are also opportunities for strategic positioning that don't require choosing sides entirely. Nations with critical mineral deposits can negotiate for downstream participation in supply chains not just exporting raw materials but hosting processing facilities and manufacturing operations. Countries with abundant renewable energy can attract data centers and computing infrastructure. Regions with strong educational systems can develop AI talent that commands global wages rather than watching their best minds emigrate.

The key insight is that passive participation, simply consuming AI applications developed elsewhere, is a path to deepening dependency. Strategic participation requires building some combination of indigenous capabilities, whether in infrastructure, talent, specialized applications, or supply chain integration that gives a nation leverage and options rather than leaving it entirely beholden to others' choices.

What Comes Next

Artificial intelligence has joined oil, semiconductors, and finance in the exclusive category of industries that determine national power and global hierarchies. The transformation happened faster than most anticipated, compressed into a few years rather than unfolding across decades. For the Americas, this creates both urgency and opportunity.

The urgency stems from how quickly positions are being locked in. The investments being made today in fabrication plants and data centers, in research facilities and talent pipelines, in alliances and technology standards will be difficult to reverse. A nation that finds itself entirely dependent on foreign AI infrastructure in 2030 will struggle to build sovereignty in 2035, because the capital requirements, technical knowledge, and ecosystem development take years to mature.

The opportunity comes from the fact that positions aren't yet entirely determined. The AI industry is still being built. Supply chains are being restructured. Standards are being negotiated. New capabilities are emerging that could shift competitive dynamics. Nations that act strategically today still have options that will be foreclosed to those who wait.

But strategy requires clear-eyed assessment of capabilities, resources, and constraints. Not every nation can or should attempt to build a complete AI stack. The question each must answer is: which elements of the AI value chain are most critical for our national interests? Where do we have genuine competitive advantages we can exploit? Which dependencies are acceptable risks, and which represent unacceptable vulnerabilities?

These aren't abstract questions for think tanks and academic papers. They're urgent strategic imperatives that will determine prosperity, security, and autonomy for hundreds of millions of people across the hemisphere.

Next week, we turn from the question of AI as strategic industry to its most immediate and politically sensitive consequence: what happens to work and workers when machines become capable of doing what once required human intelligence?

This is Part 2 of a 10-part series on The Geopolitics of Artificial Intelligence in the Americas by Core Geopolitical Insights LLC. Follow along each week as we explore how this transformative technology is reshaping power, prosperity, and security across the Western Hemisphere. | Next week: The Employment Outlook