When Ukrainian forces planned an attack on Russia’s Black Sea Fleet at Sevastopol in 2022, they encountered an obstacle beyond the usual calculations of military risk. Starlink, the satellite network on which they had come to rely, did not operate around Crimea. Extending its coverage required the agreement of SpaceX and its chief executive, Elon Musk. A military operation depended, in part, on a businessman’s consent.

Musk refused the request, later explaining that he did not want SpaceX to become “explicitly complicit in a major act of war and conflict escalation.” The episode was initially reported differently. Walter Isaacson’s biography of Musk suggested that the entrepreneur had ordered Starlink coverage around Crimea switched off as the operation approached. Isaacson subsequently acknowledged that this was incorrect: coverage had never been activated there.

The correction matters. Refusing to extend a service is different from withdrawing one already in use. But it leaves the central political question intact. A sovereign government fighting for its survival had sought the approval of a private company to use communications infrastructure for a military operation. Musk’s judgment about escalation had become a consideration in Ukraine’s ability to act.

His position is exceptional, but the dependence it exposes extends well beyond SpaceX. Governments increasingly rely on privately controlled satellite networks, cloud infrastructure, artificial intelligence systems, ports, energy networks, and commodity supply chains. Their owners remain subject to public authority. Yet control over systems essential to national security gives these companies influence that ordinary lobbying cannot explain. They help determine whether governments can carry out decisions they have already made.

Private power has shaped international politics for centuries, from the East India Company to Standard Oil. The novelty lies in the assets under corporate control and the speed with which decisions about their use can produce geopolitical consequences. A communications service can be restricted faster than a government can build an alternative. As dependence deepens, the boundary between commercial discretion and public power becomes harder to draw.

SpaceX’s relationship with Washington illustrates the problem. The company remains privately owned, yet it is deeply embedded in American space and defense infrastructure. A Reuters review published in February 2025 cited approximately $22 billion in government contracts for SpaceX, roughly $15 billion of them from NASA. Its work also includes military satellite launches, Starlink communications, and Starshield, a satellite business designed for national security applications.

The Pentagon has sought to put Ukraine’s reliance on Starlink on a firmer contractual footing. In 2023, the U.S. government began purchasing services directly for the country, reducing some of the uncertainty surrounding a system initially supplied through a mixture of private support and public arrangements. The Pentagon awarded SpaceX a contract worth about $23 million to support military operations in Ukraine through mid-2024. Contracts provided a means of defining obligations that goodwill alone could not guarantee.

States have bought weapons, ships, and aircraft from private manufacturers for centuries. In many such transactions, however, the manufacturer delivers an asset that the government subsequently operates. Satellite constellations and cloud services can create a more enduring dependence. The provider continues to run the infrastructure, maintain the software, and supply the expertise on which the customer relies. Purchasing access does not necessarily give a government the ability to operate the system itself.

AI is extending this relationship into the information systems used for military decisions. In March 2025, NATO acquired Palantir’s Maven Smart System for Allied Command Operations. According to NATO’s Communications and Information Agency, the platform supports intelligence fusion and targeting, battlespace awareness, planning, and accelerated decision-making. The acquisition took six months, making it one of the fastest in NATO’s history. That speed reflected the alliance’s determination to bring commercial technology into military use.

Later in 2025, Maven served as the primary platform for warfighting integration during Steadfast Duel, an exercise testing collective defense across NATO’s command structure. The alliance reported faster data processing and decision support across multiple domains. Software developed by a private American company was becoming part of the machinery through which allied headquarters coordinated military operations.

This does not mean that Palantir, or its co-founder Peter Thiel, determines NATO policy. The alliance sets requirements, buys the system, and retains military command. The distinction is essential. The concern is that public institutions increasingly depend on technologies whose complexity and development costs make them difficult to reproduce quickly within government. Formal authority over a decision can coexist with practical dependence on the company supplying the tools to implement it.

Physical infrastructure creates similar entanglements. In March 2025, a consortium involving BlackRock reached an agreement in principle to acquire dozens of ports from Hong Kong-based CK Hutchison. The proposed transaction included interests in the company operating Balboa and Cristóbal, at opposite ends of the Panama Canal. A commercial acquisition immediately became part of a geopolitical argument about Chinese influence over a vital trade route.

Donald Trump presented the proposed sale as supporting his effort to reduce that influence, while Beijing subjected the transaction to scrutiny. Ownership of the terminals did not amount to control of the canal, but their location made them strategically significant. Ports, data centers, telecommunications networks, and satellite constellations may belong to private companies while serving as essential parts of the infrastructure on which national power depends.

Venezuela reveals another form of this dependence. Following the political upheaval of January 2026, Washington encouraged a rapid expansion of Venezuelan oil production and exports. Some of the most consequential participants have been commodity trading companies, rather than governments or major international oil producers. Their advantage lies in their ability to connect crude supplies with buyers, financing, and transport at a speed that diplomatic announcements alone cannot achieve.

Vitol and Trafigura have become particularly important. In August, the two companies handled almost 600,000 barrels a day between them, accounting for more than half of Venezuelan crude exports. Their share of those flows gives them substantial commercial influence over the terms on which oil reaches the market. In September, Reuters reported that both traders were seeking larger discounts from PDVSA, Venezuela’s state-owned oil company, as rising shipping costs squeezed their margins.

Commodity traders do not set American or Venezuelan foreign policy. Their networks of ships, storage facilities, financing, and buyers nevertheless help determine how quickly policy becomes an actual movement of energy. Trafigura manages approximately 500 vessels globally and announced the launch of Volare Shipping in September 2026 to expand its tanker ownership. Such logistical capacity gives private companies a practical role in implementing strategies devised in government offices.

Venezuela also demonstrates the more controversial position of businessmen whose commercial interests intersect with political relationships. Alejandro Betancourt, the Venezuelan businessman associated with North American Blue Energy Partners, or NABEP, has emerged as a significant participant in the country’s changing oil sector. A Reuters investigation published in September described his transformation from the subject of investigations in several jurisdictions into a participant in a major U.S.-backed Venezuelan oil arrangement.

Reuters said that Betancourt had not been indicted in the United States and that investigations elsewhere remained ongoing. Those qualifications are indispensable. An investigation is not a finding of guilt, and political access cannot be treated as proof of wrongdoing. The broader point is that Venezuela’s reopening has created opportunities for investors and intermediaries able to operate across commercial and political networks. Their usefulness to governments can become a source of influence in its own right.

Formal diplomatic relations therefore explain only part of what happens in politically risky markets. Companies entering Venezuela must navigate PDVSA, U.S. sanctions and licensing requirements, Venezuelan regulations, financial institutions, and networks of contractors and local partners. Knowing how to connect these institutions, and gaining access to the people who make decisions within them, can be commercially valuable. A government’s objectives may depend on intermediaries whose relationships it cannot readily duplicate.

The phenomenon is not confined to Venezuela. Sanctions against Russia and Iran have increased the importance of shipping companies, traders, financial intermediaries, and opaque corporate structures capable of sustaining commerce when conventional channels become difficult to use. Western governments seek to restrict some of these networks while relying on other private companies to carry out their own strategic objectives. Corporate capabilities can facilitate public policy or frustrate it, depending on who controls them and how they are used.

It would be misleading, however, to suggest that billionaires and multinational companies are replacing states as the principal actors in international politics. Governments legislate, tax, regulate, impose sanctions, control borders, and command military force. Companies acquire geopolitical importance within frameworks that public authorities establish, and their influence can be curtailed when those authorities intervene. Corporate power is substantial, but neither unlimited nor independent of the state.

The relationship is one of growing interdependence, although the balance varies by industry and circumstance. Governments need private capital and technical expertise because some strategic systems are too expensive, specialized, or fast-changing to develop entirely within public institutions. Companies depend on governments for contracts, licenses, market access, and legal protection. In many cases, the state is also one of their largest customers. Each side holds powers the other needs.

The most consequential shift is that private companies increasingly operate the systems through which governments exercise authority. SpaceX supplies satellite launches and military communications. Palantir provides software integrated into NATO’s command structures. Infrastructure investors hold strategically located assets, while commodity traders possess the networks needed to redirect large volumes of oil. Their influence derives partly from the difficulty of replacing them when national priorities change or commercial relationships deteriorate.

History offers more extreme examples. The East India Company governed territory and maintained armed forces. Today’s technology and infrastructure companies possess nothing comparable to that formal political authority. They can nonetheless become difficult to dispense with, even for governments that retain the legal power to regulate them. The distinction between sovereignty and operational capacity helps explain why private influence can grow without a corresponding transfer of public office.

The Starlink controversy ultimately matters for reasons larger than Musk’s personality or political views. Ukraine needed a privately operated network, and its management retained discretion over where that network could be used.

Subsequent Pentagon contracts show how governments can respond by defining obligations more clearly. They also reveal the limits of that response: a more dependable contractual relationship still leaves the state reliant on infrastructure it does not operate.

As AI, satellite communications, and critical supply chains become more important to national security, governments will face harder questions about how much dependence they can accept and what alternatives they can sustain. Democratic accountability becomes difficult when decisions made inside private companies affect a government’s capacity to fulfill public commitments. States remain the dominant geopolitical actors. Their ability to act, however, increasingly rests on systems controlled by people who hold no public office and answer to a different set of obligations.