By Vanguard with the work of Michael Porter, Pankaj Ghemawat, Ian Bremmer, Mariana Mazzucato, and Daron Acemoglu.

Global capitalism is being reorganized around a new set of constraints. For decades, the dominant logic of globalization was efficiency. Capital moved toward lower costs. Supply chains stretched across borders. Production specialized by region. Technology firms scaled on global platforms. Energy markets supported industrial expansion, and geopolitical risk was often treated as a disruption to be managed rather than a permanent feature of strategy.

That era has not disappeared, but it has weakened. In 2026, capitalism is being reshaped by three transitions at once: the acceleration of artificial intelligence, the reconfiguration of energy systems, and the geopolitical realignment of supply chains, technology, and industrial policy. These transitions are not separate. They are converging into a new economic order in which compute, electricity, semiconductors, data, minerals, talent, and national security increasingly determine competitive advantage.

The result is not deglobalization in the simple sense. The world remains deeply connected. Trade continues. Capital still crosses borders. Companies still depend on global customers, suppliers, and innovation networks. But the assumptions behind global capitalism are changing. Efficiency is no longer enough. Resilience, sovereignty, security, energy access, and political alignment now shape where companies invest, how they source, which technologies they deploy, and what risks they accept.

The central question for executives is no longer whether the global economy will remain open or fragment completely. It is how to compete in a world that is open in some domains, restricted in others, and increasingly organized around rival systems of power.

The Triple Transition

The first transition is AI acceleration. Artificial intelligence is moving from software capability to strategic infrastructure. The competitive race is no longer only about who has the best models. It is about who controls the stack beneath them: advanced chips, data centers, energy, cooling systems, cloud platforms, data access, talent pipelines, model governance, and deployment infrastructure. AI is becoming an industrial system.

This changes the economic map. Countries and companies with access to compute, energy, and advanced semiconductor supply chains will have different possibilities than those without them. AI capability is not evenly distributed. It concentrates where capital, infrastructure, talent, and energy can be assembled at scale. That concentration creates productivity opportunity, but also dependency.

The second transition is energy. AI makes electricity a strategic input to intelligence. Data centers require enormous and reliable power. The energy transition requires new grids, storage, renewables, nuclear investment, transmission, critical minerals, and policy coordination. Industrial competitiveness increasingly depends on whether a region can provide clean, affordable, stable energy at scale. The old distinction between digital strategy and energy strategy is collapsing.

The third transition is geopolitical realignment. Governments are using industrial policy, export controls, sanctions, tariffs, investment screening, procurement rules, and data regulations to shape economic outcomes. Supply chains are being redesigned not only for cost, but for political reliability. Companies are expected to understand not only suppliers and customers, but the strategic intent of states.

Together, these transitions are reinventing capitalism from the inside. The firm is no longer merely a market actor optimizing across a global field of opportunities. It is increasingly a participant in systems of national strategy, technological competition, and resource constraint.

From Efficiency to Sovereignty

Sovereignty has become one of the defining words of the new economic era. Governments speak of energy sovereignty, data sovereignty, technology sovereignty, food sovereignty, and now AI sovereignty. The idea is simple: critical capabilities should not be entirely dependent on foreign systems that may become unavailable, restricted, or politically compromised.

For companies, sovereignty creates both opportunity and complexity. Governments are investing in domestic capacity, regional supply chains, cloud infrastructure, chip production, clean energy, and strategic industries. These investments can create new markets and incentives. But they also create fragmentation. A company may face one data regime in Europe, another in China, another in the United States, and still another in emerging markets seeking their own digital autonomy.

Sovereign AI is a particularly important example. It is not enough for a country to declare that it wants domestic control over AI. Practical sovereignty requires infrastructure: chips, data centers, energy, network capacity, cloud capability, cybersecurity, talent, and governance. Without those assets, sovereignty becomes rhetoric.

Businesses must recognize that sovereign AI will not be merely a government project. It will affect enterprise architecture. Multinationals may need different model deployments, data-storage arrangements, vendor relationships, and compliance structures across regions. They may need to decide which AI systems can operate globally and which must be localized. They may face pressure to use domestic clouds, domestic models, or approved infrastructure providers.

The strategic question becomes: where should a company preserve global scale, and where must it adapt to sovereign constraints?

The New Variants of Capitalism

The reinvention of global capitalism will not produce one model. It will produce competing variants.

The United States model remains innovation-led, capital-intensive, and market-oriented, but with increasing strategic state intervention. American capitalism still benefits from deep capital markets, world-leading technology firms, elite universities, entrepreneurial culture, and a powerful cloud and AI ecosystem. Yet it is no longer purely laissez-faire. Industrial policy, export controls, semiconductor incentives, defense-linked innovation, and national-security screening now play a larger role. The U.S. model combines private-sector dynamism with strategic intervention in sectors considered critical to future power.

The European model emphasizes regulation, rights, sustainability, competition policy, and social legitimacy. Europe seeks to shape digital capitalism through privacy rules, AI regulation, sustainability reporting, and competition enforcement. Its strength is normative power: the ability to set standards that influence global practice. Its challenge is speed, scale, energy cost, fragmented capital markets, and relative weakness in frontier AI infrastructure. Europe’s question is whether a rules-based model can also generate enough industrial capacity to remain strategically relevant.

The Chinese model is state-directed, industrially ambitious, and sovereignty-focused. China combines large-scale manufacturing, infrastructure execution, state planning, domestic platforms, and strategic technology policy. Its system can mobilize resources quickly and align firms with national priorities. But it faces constraints from demographics, capital allocation inefficiencies, global trust concerns, export controls, and geopolitical tension. China’s version of capitalism emphasizes control, scale, and industrial depth, but must manage the cost of reduced openness.

Other regions are not simply passive. The Gulf states are turning energy wealth into AI and infrastructure ambitions. India is positioning itself around digital public infrastructure, services, manufacturing, and strategic autonomy. Southeast Asia is benefiting from supply-chain diversification. Africa and Latin America face risks of being left behind in AI infrastructure, but also possess critical minerals, demographic potential, and opportunities for leapfrogging if investment, governance, and energy access improve.

The future of capitalism will therefore be plural. Executives must stop assuming that one global operating model will fit every market. Strategy will require fluency across variants of capitalism.

Supply Chains as Strategic Architecture

Supply chains used to be discussed primarily in terms of efficiency and risk. They are now strategic architecture. The location of production, the ownership of suppliers, the jurisdiction of data, the origin of chips, the source of energy, and the political alignment of partners all matter.

This does not mean every company should reshore everything. Full reshoring is often too expensive, impractical, or counterproductive. But companies must become more sophisticated about dependency. They need to know which inputs are critical, which suppliers are replaceable, which regions create geopolitical exposure, and which capabilities must remain under closer control.

AI makes this more urgent. Advanced chips, cloud services, data centers, rare minerals, energy infrastructure, and specialized talent are now part of the supply chain for intelligence. A company that treats AI only as software may miss the physical and geopolitical dependencies beneath it. If its AI strategy depends on a narrow set of chip suppliers, cloud providers, energy markets, or regulatory permissions, its competitive position is more fragile than it appears.

Supply-chain strategy should therefore include four layers. The first is visibility: knowing where inputs, data, and dependencies actually reside. The second is optionality: building alternative suppliers, regions, and technical pathways. The third is control: identifying what must be owned, governed, or protected internally. The fourth is collaboration: working with governments, suppliers, and industry partners where no single firm can solve systemic risk alone.

The strongest companies will not optimize only for cost. They will optimize for continuity, adaptability, trust, and strategic relevance.

Investment in a Fragmented World

Capital allocation is also changing. Investors and executives must now evaluate not only market size and return potential, but geopolitical exposure, energy availability, regulatory alignment, infrastructure capacity, and sovereignty risk.

A data center investment is no longer merely a real-estate or technology decision. It is an energy decision, a water decision, a grid decision, a permitting decision, and a geopolitical decision. A semiconductor investment is not merely a manufacturing decision. It is a national-security decision, a supply-chain decision, and a diplomatic decision. A clean-energy investment is not merely an environmental decision. It is an industrial competitiveness decision.

This makes capital allocation more complex but also more strategic. Companies that understand the new constraints can identify where future bottlenecks will create value. Power availability, grid modernization, cooling technology, chip packaging, critical minerals, cybersecurity, energy storage, trusted cloud infrastructure, AI governance tools, and supply-chain verification may all become strategic growth areas.

At the same time, firms must avoid chasing government subsidies without strategic coherence. Industrial policy can create attractive incentives, but it can also distort investment if companies pursue subsidies rather than durable advantage. The question should not be whether public support is available. It should be whether the investment strengthens the company’s long-term position under plausible geopolitical and technological scenarios.

In the new capitalism, capital must be patient enough for infrastructure and agile enough for volatility.

Cross-Border Collaboration Under Constraint

The paradox of the coming era is that global challenges require collaboration even as geopolitical trust declines. AI safety, climate transition, semiconductor supply, energy security, cyber risk, pandemic readiness, and financial stability cannot be solved entirely within national borders. Yet the political willingness to depend on rivals is shrinking.

Companies will have to operate in this tension. They will be asked to collaborate globally while complying with restrictions locally. They will need partnerships in markets where political relations are strained. They will need research networks that cross borders, but with stronger controls around intellectual property and dual-use technology. They will need supply chains that remain global, but less naive about strategic dependency.

Cross-border collaboration will therefore become more selective. The old model was broad integration. The new model is conditional integration. Companies will collaborate where trust, rules, and mutual interest are sufficient. They will restrict collaboration where national-security, data, or human-rights risks are too high. They will build governance systems to decide which is which.

This requires a more mature form of globalization. The goal is not maximum openness at all costs. Nor is it isolation. It is governed interdependence.

Four Scenarios for Global Capitalism

Executives should prepare for four plausible scenarios.

The first is managed fragmentation. In this scenario, the world divides into partially aligned blocs, but trade and investment continue through controlled channels. Companies regionalize some operations, localize sensitive data, diversify suppliers, and adapt to multiple regulatory regimes. This is difficult, but manageable for firms with strong geopolitical intelligence and modular operating models.

The second is technology bifurcation. In this scenario, AI, cloud, chips, cybersecurity, and data standards split more sharply across U.S.-aligned, China-aligned, and regionally autonomous systems. Companies must choose technology stacks by market and accept reduced interoperability. The winners are those that design architectures flexible enough to operate across regimes without losing control.

The third is energy-constrained AI. In this scenario, AI demand grows faster than power systems can support. Data-center expansion faces grid bottlenecks, permitting delays, water constraints, and political resistance. Companies with access to reliable energy and efficient infrastructure gain advantage. AI strategy becomes inseparable from energy strategy.

The fourth is renewed pragmatic integration. In this scenario, the costs of fragmentation become too high, prompting selective cooperation around climate, AI safety, financial stability, and supply-chain resilience. Competition remains, but governments and firms rebuild limited trust where mutual dependence is unavoidable. Companies that maintained relationships across borders benefit.

No executive can know which scenario will dominate. The purpose of scenario planning is not prediction. It is preparation.

The Strategic Imperatives for Executives

Leaders should begin by mapping strategic dependencies. This means identifying where the company depends on specific countries, suppliers, technologies, data regimes, energy sources, infrastructure providers, and regulatory permissions. Dependency is not automatically bad. Invisible dependency is.

Second, companies should build modularity into operations. A modular organization can adapt products, technology stacks, data systems, supply chains, and governance structures across regions without reinventing the whole enterprise. Modularity is the operating model for a fragmented world.

Third, executives should integrate energy into digital strategy. AI ambitions must be tested against power availability, cost, carbon intensity, grid resilience, and data-center strategy. A company that has an AI roadmap without an energy view has an incomplete strategy.

Fourth, leaders should develop geopolitical literacy below the board level. Geopolitical risk cannot remain a specialist function. Business-unit leaders, supply-chain managers, technology teams, legal departments, and capital-allocation committees need the ability to recognize how policy shifts affect operations.

Fifth, companies should preserve optionality. This includes supplier diversification, multi-cloud strategies, local partnerships, alternative logistics routes, flexible manufacturing, and internal capabilities in areas too critical to outsource entirely. Optionality has a cost, but so does fragility.

Finally, executives should maintain a principled approach to markets. In a multipolar world, companies will face pressure to compromise values for access. They will need clear standards around human rights, data use, corruption, surveillance, environmental claims, and security-sensitive partnerships. Capitalism without principles becomes opportunism. Opportunism becomes risk.

The Real Reinvention

The reinvention of global capitalism will not be announced by a single summit or captured in a single policy. It is already happening through thousands of decisions: where to build data centers, where to source chips, where to locate factories, where to store data, which markets to enter, which suppliers to trust, which technologies to deploy, and which governments to align with.

The companies that succeed will not simply be global or local. They will be adaptive. They will combine global intelligence with regional flexibility. They will treat AI as infrastructure, energy as strategy, and geopolitics as a core management discipline. They will understand that supply chains are no longer back-office systems but expressions of strategic posture.

The companies that fail will cling to old assumptions. They will optimize for cost in a world that demands resilience. They will pursue AI without understanding energy. They will enter markets without understanding political alignment. They will treat sovereignty as a compliance issue rather than a strategic force. They will discover that the world has changed only after their operating model breaks.

Capitalism is not disappearing. It is being reorganized. The next version will be more industrial, more digital, more energy-constrained, more politicized, and more regionally differentiated. It will reward companies that can learn faster than institutions fragment.

In the old global economy, advantage belonged to firms that scaled efficiently across borders. In the new one, advantage will belong to firms that can adapt intelligently across systems.

That is the new executive task: to build organizations capable of competing where capitalism is no longer one game, but many.