Why decentralization is the future of internet governance

A registry can keep records for networks across a continent. That does not give it a mandate to speak for everyone on that continent.

This distinction sits at the centre of Lu Heng’s argument about Internet governance. An administrative task—coordinating names or numbers—can grow into a claim to political authority. The people affected may never have appointed the institution, yet their networks depend on its decisions.

Where the authority claim goes wrong

A membership organization can speak for its members within the authority they give it. An advisory group can offer advice. Neither automatically represents all Internet users in the places it serves.

In Note 54, Lu Heng challenges the leap from a service area to a political mandate. Drawing a regional boundary, assigning committee seats and holding meetings does not mean the wider population has delegated power to that institution.

The practical question follows: when an administrator acts beyond its record-keeping role, what can a network do? If it cannot leave without putting recognition of its resources at risk, consultation alone cannot balance that power.

The people carrying the risk need a real choice

Imagine an operator that has built a network around an address range. It pays for equipment, employs engineers and answers to customers when service fails. A registry maintains the records, but it does not take over those obligations when its own decisions cause uncertainty.

Lu Heng’s position is that coordination should serve the resource holder. It should not become a superior claim over the resources or a licence to govern the operator’s customers. Note 28 explains why he opposes registries turning into enforcers: global technical coordination should not become a private system of punishment.

Law still matters. Courts and public authorities act within their jurisdictions. The issue is whether a private registration role should acquire a separate, cross-border coercive power simply because networks need their records recognized.

The solution changes who can exercise power

Decentralization must reach the registry itself. Resource holders need records that others can verify independently, a choice of service providers and a way to keep operating when one institution fails.

A distributed registry is intended to make accepted records depend on shared verification rather than one administrator’s permission. Changes require valid proofs and common rules. No single organization should be able to erase another network’s standing by itself.

That is why replacing today’s administrator with a new global authority misses the point. The name on the office door changes; the inability to leave remains. Lu Heng argues in Note 8 that a decentralized registry is the only viable path out of this structural dependence.

Why waiting makes the problem harder

The need for dependable connectivity does not pause during an institutional crisis. Networks still have customers, bills and services to deliver. A system that can change administrators while preserving trusted records is better prepared for that reality.

The work has to begin before failure forces the issue: establish how records are verified, how conflicting claims are resolved and how existing networks can adopt the new system without losing continuity. The proposal has to earn operational acceptance; declaring a ledger “decentralized” is not enough.

In Note 37, Lu Heng describes NRS as an advocacy group pushing for this change. Its role is to argue for the direction, not to sell a finished registry product. The destination is number governance that no single institution can hold hostage.

A simple way to judge a reform

Ask three questions. Can an operator leave? Can other networks still verify its resource records? Can services continue if the administrator fails?

If a reform only adds meetings but leaves those questions unanswered, it has not removed the dependency. To explore the argument behind this standard, read Lu Heng’s Note 54 on who gets to speak for a continent or “the end user”.