Generated by Codex with GPT 5.6 Sol XHigh

The United States does not dominate every foreign government contract. It dominates many of the contracts that would be hardest to replace. Cloud platforms, enterprise software, fighter jets, air-defence systems and hospital records sit deep inside the machinery of the state. As America’s alliances become less predictable, that concentration is turning an old commercial advantage into a strategic vulnerability for its partners.

The article opens in Schleswig-Holstein, where the German state government has moved about 30,000 civil servants from Microsoft’s collaboration and productivity tools to open-source alternatives. Other European cities and governments are experimenting with similar changes. Their concern is no longer merely cost or privacy. They fear that Washington could restrict access to essential American technology during a political dispute, just as it has used sanctions and export controls against adversaries.

Recent events have made the possibility feel less theoretical. The chief prosecutor of the International Criminal Court reportedly lost access to his Microsoft email after the Trump administration sanctioned the court, though Microsoft says it did not cut the service. The United States also temporarily forced Anthropic to withhold its newest artificial-intelligence models from foreign users and twice paused arms deliveries to Ukraine. Each episode reminded allies that a product supplied by an American company can come with exposure to American policy.

Small share, critical role

American firms win only a modest portion of public contracts abroad. The Economist estimates that foreign governments contributed roughly \$500bn of the \$25trn in sales generated by listed American companies last year. That equals about 7% of government procurement in OECD countries outside the United States. Dependence also varies: American companies won around 6% of contracts in Australia and 4% in Britain, compared with 2% in France and 1% in Germany.

Those averages understate the real leverage because the contracts are concentrated in critical sectors. Information technology and defence account for about two-fifths of American companies’ foreign-government business, while pharmaceuticals and medical equipment make up another substantial share. Alphabet, Amazon and Microsoft control roughly two-thirds of the global cloud-computing market. In many categories of enterprise software, American firms occupy most of the leading positions. Governments seeking the most capable air-defence systems have little practical alternative to the Patriot, made by Lockheed Martin and RTX.

This is not ordinary dependence on imports. These systems underpin communications, military operations, health care and public administration. Replacing them can require new equipment, retrained staff, redesigned workflows and years of transition. A country that abandons the F-35, for example, must change far more than its aircraft: pilots and maintenance crews need retraining, weapons inventories need replacing and an entire operational system must be rebuilt.

Sovereignty has a price

European governments are trying to cultivate domestic substitutes. France wants to replace Microsoft Teams in its national administration. Spain has cancelled an order for F-35s. The European Commission has proposed moving sensitive government data to European cloud providers. Local administrations are adopting open-source office software and regional hosting companies.

Yet the alternatives often sacrifice capability or scale. LibreOffice lacks Microsoft’s fast-growing AI features. France’s OVHcloud produces about one-hundredth the revenue of Amazon Web Services, limiting how much it can spend on price competition and innovation. European combat aircraft may be less advanced than their American counterparts. Even apparently local systems can rest on American foundations: OVHcloud runs much of its infrastructure on software from VMware, which belongs to Broadcom.

Political retaliation creates a second cost. A government that visibly rejects American suppliers may antagonise Washington and make the alliance itself less reliable. Denmark’s decision to buy Boeing maritime-patrol aircraft instead of European alternatives may reflect not only technical judgment but also a desire to avoid another dispute with the United States. Efforts to reduce dependence can therefore provoke the very geopolitical risk they are meant to contain.

From independence to mutual leverage

The article’s strongest answer is not complete technological self-sufficiency, which is both unrealistic and expensive. Modern supply chains are too interdependent. America’s federal government relies heavily on software from Germany’s SAP. Every F-35 contains a rear fuselage made by Britain’s BAE Systems. The most advanced cloud-computing chips are manufactured by Taiwan’s TSMC using machines supplied by the Netherlands’ ASML.

That suggests a more practical strategy for America’s allies: preserve credible alternatives where possible, but also deepen the areas in which the United States depends on them. Open-source software, European cloud providers and domestic defence production can reduce the most dangerous single points of failure. At the same time, indispensable components and specialised expertise give allies bargaining power inside an interconnected system.

The goal, then, is not to purge American companies from government supply chains. Their products are often better, and replacing them indiscriminately would waste money and weaken public services. The goal is to distinguish ordinary commercial convenience from strategic dependence, build fallback options for essential functions and make one-sided coercion costly. In a less dependable alliance, resilience comes not from isolation but from ensuring that dependence runs both ways.