Who really owns the Internet?

Nobody owns the Internet as a whole. But that does not mean the networks, equipment and resources that make it work belong to nobody—or to everyone collectively.

Your connection crosses systems run by different organizations. A provider operates the access network. Other operators carry traffic between networks. A service runs the application you are using. Each has its own assets, responsibilities and points of control.

The more useful question is therefore: which part are we talking about, who depends on it, and who can change the rules? That question leads directly to Lu Heng’s argument about Internet number resources.

Separate ownership, operation and coordination

A company can own a cable or a data centre. A network operator can run routers and decide how it connects to other networks. A platform can set rules for its own service. None of those roles amounts to owning the whole Internet.

Technical coordination is another role. The IETF develops voluntary standards that help independently built systems work together. IANA maintains shared registries for identifiers, including domain names, number resources and protocol parameters. Regional registries provide number-resource services. These functions matter, but writing a standard or maintaining a record is different from owning the systems that use it.

Governments also affect networks and services through public authority within their jurisdictions. That power does not make any one government the owner of the global network.

What about IP addresses?

An IP address is a numerical identifier used in communication between networks. Consistent records help networks distinguish valid resource claims from conflicting ones. This is why the registry matters.

In Note 48, Lu Heng traces the change from a simple technical identifier to something around which operators invest, trade, finance and build businesses. An address range can become deeply embedded in a working network. Replacing it may mean much more than editing a line in a database.

His position is that the operator holding and using the resource, investing in it and bearing its risks is the substantive asset holder. The registry provides a recording and coordination service. Keeping the record does not give the administrator a superior ownership claim.

A service region does not make addresses a continent’s property

Regional registries serve different parts of the world. That arrangement organizes a service; it does not turn every resource recorded there into the political property of the region.

Lu Heng makes this distinction in Note 53. Networks build the businesses and carry the losses when continuity or recognition is disrupted. He argues that a registry cannot replace those realities with a claim that it speaks for a continent or owns resources on the continent’s behalf.

The same distinction prevents two common mistakes: treating the failure of one registry as a judgment on an entire region, and treating criticism of that registry as an attack on the region’s people. An institution and the people inside its service area are not the same thing.

No single owner does not mean no concentrated power

You can depend heavily on something without its provider owning your whole business. A platform may control access to an audience. A network provider may be difficult to replace. A registry may hold records other networks rely on.

These are different kinds of dependence. Moving to a different application does not automatically change how your addresses are registered. Changing the registry does not automatically give you control over a platform’s recommendation system.

Lu Heng focuses on the registration layer because decisions there can affect networks that have already invested in the resources. In Note 49, he explains how that influence depends on other participants accepting the registry as their reference. The practical power is real, even though the registry does not own every cable or operate every router.

The alternative: keep reliable records without an irreplaceable administrator

The proposed answer is not to hand the entire Internet to a new owner. It is to make the coordination service replaceable while keeping its records dependable.

Resource holders need proofs that others can verify, consistent transfer records and a way to retain recognition when a service provider changes. The common system must prevent conflicting claims. It should also let a functioning network survive the failure of an institution above it.

This is the direction set out in Lu Heng’s Note 72: preserve unique identifiers, accurate records and operational continuity; give operators a meaningful ability to leave. Decentralization has to deliver those capabilities in practice. Merely distributing copies of a database does not establish them.

Why does this matter now?

When more investment and customer service depend on an address range, uncertainty over its recognition becomes more costly. Waiting until an institution fails leaves operators trying to protect live services while searching for an alternative.

The right time to establish portable records and tested continuity arrangements is before that emergency. The aim is not a world without coordination. It is a world where coordination does not let one administrator hold other people’s networks hostage.

So, who owns the Internet? There is no single owner. Its many participants need common rules to cooperate, but those rules should serve the networks that make the Internet exist. Read Lu Heng’s Note 53 for the argument about why recording a resource must not become a claim to rule it.