Privatized Networks Don’t Just Connect the World. They Can Steer It.

For builders and investors, the tempting assumption is that satellite networks, submarine cables, cloud regions, and messaging platforms are mostly plumbing: expensive, technical, but politically neutral. The evidence here points the other way. Once a few private operators control the channels that carry warfighting data, diplomatic messages, or national internet traffic, they acquire leverage that states cannot always replace quickly, if at all.

That does not mean every private network operator is a geopolitical actor in the same sense as a state. It does mean the line between commercial infrastructure and sovereign function is now blurry enough to matter in product strategy, procurement, and national security planning.

The core shift: from service provider to gatekeeper

The best shorthand for the shift is that private firms now sit between governments and the communications systems they depend on. In some cases that influence is formalized through infrastructure ownership. In others, it shows up as operational discretion: who gets service, where coverage is allowed, which traffic gets prioritized, and which regions are effectively deplatformed.

A recent analysis of Starlink puts the point plainly: “The twenty-first century’s real geopolitical contest is no longer only between states, but between states and the corporations that command the world’s digital infrastructure.” ¹

That claim is broad, but the mechanism is concrete. Starlink’s role in Ukraine shows how a private network can become part of battlefield logistics, reconnaissance, and command-and-control. By early 2025, Ukraine was using at least 47,000 Starlink terminals, and the network became critical for decentralized fire-support and drone operations. ² Another 1 Minute Signal coverage of Business Insider highlights that Ukraine’s smaller defense stack has adapted around Starlink-integrated platforms such as the Peroon Max, which uses Starlink to reduce dependence on vulnerable radio frequencies and GPS. ³

"When a private supplier can decide which operations a front-line state is permitted to conduct based on personal intuition, the relationship has ceased to be commercial. It is a delegation of sovereignty, a strategic function exercised by an unaccountable executive."

— Foreign Policy ²

That is the heart of the issue for founders: once your infrastructure becomes critical, your governance model becomes geopolitical whether you planned for it or not.

Why Starlink is the clearest warning case

Starlink is not just a successful product. It is also a prototype for what happens when a private platform becomes too embedded to swap out easily.

Foreign Policy’s analysis notes that Starlink has already been used in ways that affected the tactical balance in active conflict, including the 2022 restriction near Russian-occupied Crimea and later authentication controls to prevent misuse in occupied territories. ² The paper’s framing of “private discretion” matters because it describes a power that is not exercised through treaties or formal military command, but through terms of service and operator judgment.

That creates a strange asymmetry. States still carry the accountability for outcomes. Private providers often hold the operational keys.

This is also why Musk’s own comments about a whitelist of approved Starlink terminals matter beyond Ukraine. 1 Minute Signal coverage of The Economist reports that he worked with the government to create that whitelist to balance operational necessity with regional security risks. ⁴ In practice, that means access policy can become a live instrument of conflict management.

The risk is not merely that a network can be turned off. It is that a network’s default commercial logic can be overruled by ad hoc geopolitical judgment.

"Musk explains he worked with the government to create a whitelist of approved Starlink terminals to balance operational necessity with regional security risks."

— 1 Minute Signal coverage of The Economist ⁴

Subsea cables and peering: the quieter version of the same problem

Starlink gets the headlines because satellites are visible and conflict is obvious. But the deeper structural dependency still runs through undersea cables and internet peering.

Submarine cables carry over 99% of global internet traffic, and their private, cross-jurisdictional ownership makes digital sovereignty difficult to preserve in practice. ⁵ A separate SSRN paper makes the same point more bluntly: submarine cables “curtail” a state’s ability to regulate the infrastructure it relies on, reduce data security, and threaten telecom service continuity. ⁶

The governance problem is not only ownership. It is topology.

The Internet Society warns that centralized peering makes it easier for a state to isolate an entire population and vital services, because traffic can be funneled through a few large transit paths. ⁷ In other words, the more infrastructure gets optimized for efficiency and scale, the more it can resemble a switchboard for coercion.

That matters for builders designing carrier strategy, cloud placement, or enterprise network architecture. “Resilience” is not a marketing adjective here. It is a political property.

The same pattern appears in the legal and commercial architecture around cables. A Harvard International Law Journal piece traces how international legal regimes and state-corporate partnerships shaped today’s cable networks, while also showing that cables remain a site of continuing contestation among states, developers, and corporations. ⁸ A more recent report on the ownership shift adds that hyperscalers like Meta, Alphabet, Microsoft, and Amazon have become central subsea actors, while governments increasingly view these systems as national security assets. ⁹

Private infrastructure is becoming a diplomatic instrument

The geopolitical influence is not limited to war. It also shows up in where companies place infrastructure and how that placement aligns countries into particular technical and legal ecosystems.

A 2026 analysis from the Center for Cyber Diplomacy and International Security argues that “The physical placement of a data centre is now a diplomatic act.” ¹⁰ That sounds abstract until you look at the numbers: the five largest hyperscalers are projected to spend $600 billion globally in 2026, with Microsoft alone committing $80 billion to AI infrastructure. ¹⁰ Those investments are not just CapEx. They are bargaining power.

The same logic applies to SpaceX’s proposed move to push AI infrastructure into orbit. 1 Minute Signal coverage of The Economist says SpaceX is banking on moving AI data centers into orbit to leverage solar power and circumvent terrestrial regulatory friction, a strategy the transcript itself describes as technologically unproven. ¹¹ Ben Thompson’s commentary, summarized by 1 Minute Signal, makes the broader strategic bet even more explicit: space-based compute is partly an answer to terrestrial zoning friction and permitting constraints. ¹²

For builders, the lesson is not that orbital data centers are imminent. It is that infrastructure location itself can become a route around national regulation, local opposition, and even market competition. That is a geopolitical capability, not just an engineering choice.

"SpaceX succeeds by using massive scale to shift market rules"

— 1 Minute Signal coverage of Stratechery ¹²

The destabilizing edge: private networks change escalation dynamics

The most serious argument against treating privatized global networks as neutral utilities comes from security research on satellite mega-constellations.

The European Journal of International Security notes that mega-constellations can reduce satellite vulnerability and lower first-strike incentives when states have symmetric dependence on space. But under asymmetric dependence, they can do the opposite: create incentives for destructive counterspace attacks, including nuclear ASAT options, because those may be the only reliable way to disable a dominant constellation. ¹³

That is the paradox. More distributed commercial satellites can make some systems harder to knock out, while also making them so strategically valuable that they raise the stakes of conflict.

This is why Starlink in Ukraine is not a one-off story. It is a precedent. Once a private constellation becomes embedded in military and civil functions, adversaries have to plan around it, allies have to negotiate access to it, and regulators have to ask what happens when an executive’s judgment becomes a battlefield variable. ^{1, 2}

The same logic extends to other private digital systems. One Carnegie Endowment analysis describes a world where “states and the corporations that command the world’s digital infrastructure” are the main geopolitical contest, while another paper on corporate quasi-sovereignty says firms can now substitute, bypass, or preempt state power. ^{1, 14} That is not a theoretical flourish. It is a description of the institutional gap current law has not closed.

What this means for AI builders and investors

If you build AI systems, cloud tooling, connectivity products, or defense-adjacent infrastructure, there are three practical conclusions.

First, ownership structure is now a product risk. If your stack depends on a handful of operators, your uptime, compliance, and customer trust can become subject to non-market decisions. ^{2, 6}

Second, geography still matters. Hyperscaler regions, cable landings, and satellite coverage determine which laws apply, which partners are tolerated, and which governments can pressure whom in a crisis. ^{10, 15}

Third, sovereignty is becoming an infrastructure attribute. The Economist coverage of European AI strategy frames sovereignty as “owning the keys” to the infrastructure, and says that is mandatory for defense and continuity of core state services. ¹⁶ That logic increasingly applies to communications as much as compute.

"Sovereign control, defined as 'owning the keys' to the infrastructure, is mandatory for defense and the guaranteed continuity of core state services, regardless of how other global technologies perform."

— 1 Minute Signal coverage of The Economist ¹⁶

There is a reason this keeps showing up in war, sanctions, diplomacy, and content moderation instead of just networking conferences. Private communications infrastructure now sits in the path of state power. That makes its operators more than vendors, and its design choices more than technical preferences.

The opportunity for builders is real: resilient, redundant, locally governed systems are valuable. But the hidden cost is equally real: once a network becomes indispensable, its control surface becomes a political asset. The market may call that scale. States will call it leverage.