CORPORATE WORLD ORDER

The next world order is taking shape while Washington and Beijing compete for economic and military advantage. Both governments treat artificial intelligence as a strategic priority. Both depend increasingly on corporations controlling chips, computing facilities, and software. A country can write its own laws and still find that essential parts of its economy operate at the discretion of executives whom its citizens cannot remove.

The danger has a documented history. The Dutch East India Company received authority to wage war and govern territory. Britain’s East India Company ruled much of India until the British government took direct control in 1858. In Guatemala, United Fruit’s interests converged with American support for the overthrow of an elected government in 1954. Governments empowered these companies, protected their interests, and left populations to absorb the consequences.

Today’s companies need no colonial charters to acquire leverage. They supply the infrastructure that makes modern administration, communication, commerce, and defense possible. Governments purchase their services, integrate their products, and build procedures around their platforms. Once dependence grows, changing suppliers can be expensive, disruptive, or impossible in the short term. Those contracts affect the public beyond the companies signing them.

Advanced AI requires specialized processors, enormous computing facilities, reliable electricity, and financing beyond the reach of most prospective competitors. NVIDIA supplies processors essential to much of the industry. A handful of cloud companies operate the facilities. Developers seeking to challenge the largest firms often must rent computing capacity from the same businesses that already dominate the market. Concentration is built into the cost of entry.

The Federal Trade Commission documented the problem in January 2025. Its examination of partnerships between major cloud providers and AI developers identified contractual rights, access to sensitive information, and agreements committing developers to spend substantial portions of investment funding on their investors’ own cloud services. An investor can become a supplier, a commercial gatekeeper, and a potential competitor all at once. The implications extend well beyond the companies signing the contracts.

In September 2025, NVIDIA and OpenAI announced plans for a partnership involving at least ten gigawatts of computing infrastructure and potential investment reaching $100 billion as deployment proceeded. That was a conditional proposal, not proof of money already transferred. The size of the announcement matters because access to financing and equipment now determines which companies can attempt to develop the most capable systems.

The Trump administration is accelerating this arrangement through federal policy. Its July 2025 AI Action Plan called for faster data-center construction, fewer regulatory obstacles, and American technological dominance. An executive order directed federal agencies to identify financial support and ease permitting for qualifying infrastructure. China is investing heavily in the same technologies. Competition does not relieve American officials of responsibility for deciding how much public support, electricity, and regulatory accommodation these private projects receive.

The electricity demands alone deserve more attention than they get. The International Energy Agency projects that global data-center consumption will rise from approximately 485 terawatt-hours in 2025 to 950 in 2030. Those figures concern all data centers, with AI a major contributor. Power generation and transmission cannot materialize simply because investors announce another facility. Communities must contend with development, water requirements, grid capacity, and decisions about who pays for necessary upgrades.

Developers promise jobs, investment, and growth. Public officials need to measure those promises against enforceable commitments and actual costs. A tax incentive is public revenue forgone. A subsidized grid connection allocates resources. An expedited permit carries consequences for surrounding communities. Residents have every reason to scrutinize those arrangements before corporate demands become the justification for each public concession.

The labor question is just as immediate. The International Monetary Fund estimates that AI could affect roughly 40 percent of jobs worldwide and 60 percent in advanced economies. Some work will become more productive; some positions may be displaced. Those estimates establish exposure, not a forecast of mass unemployment. They nevertheless demand an answer from governments that already struggle to protect workers through previous waves of economic change.

Productivity does not distribute its own rewards. Owners decide how businesses use technology, employers set wages within legal and market constraints, and legislators write the tax and labor rules. Greater output can support higher wages, shorter hours, and stronger public services. It can also enrich shareholders while displaced workers lose bargaining power. The machines do not choose between those outcomes. People holding economic and political power do.

In Ukraine, this dependence reaches the battlefield. SpaceX’s Starlink service supports communications during Russia’s invasion, including military operations. Decisions about network access and coverage can carry consequences for a sovereign government’s armed forces. That is considerable practical influence for a commercial supplier. Other governments face related questions whenever critical defense capabilities rely on software, networks, or equipment with few workable substitutes.

Governments still possess the legal power to intervene. Antitrust enforcement, procurement conditions, taxation, and European technology regulation provide tools to challenge concentrated corporate power. The obstacles are political willingness, institutional expertise, and the cost of acting after dependency becomes entrenched. A government cannot supervise technology it lacks the authority or ability to examine. It cannot ensure continuity when a provider can withdraw an essential service and no alternative is ready.

There is no need to predict conscious machines taking control of civilization. Human executives and public officials make the decisions demanding accountability right now. They determine who obtains computing capacity, which communities absorb infrastructure costs, how public contracts are written, and whether the gains from automation are widely shared. Political power operates through those choices, even when no corporation holds formal government office.

America’s contest with China will not settle who actually governs indispensable technology. That depends on procurement, regulation, ownership, public investment, and the terms governments are willing to accept. History establishes that states can create immense private power and then struggle with its consequences. The immediate danger is a government that remains sovereign in law while becoming so dependent on a handful of corporations that enforcing the public’s democratic authority carries costs its leaders are unwilling to bear.