In my recent note, When Registry Power Detaches From Liability, It Detaches From Reality, I argued that the modern registry problem begins with a structural asymmetry. Institutions that can materially affect transferability, recognition, continuity, and therefore real economic value still operate under liability structures designed for a much earlier and much smaller world. IPv4 is now scarce, transferable, priced, litigated, financeable, and operationally indispensable, yet the legal form of the registry layer still behaves as though it were merely administering a low-value clerical record. That is the first fracture. Power has expanded; accountability has not.

But that liability fracture does not stand alone. It sits on top of an older design error that is now becoming impossible to ignore. The Internet was not born as five political spaces. It was born as one network. The move toward Regional Internet Registries was not the discovery of a deep truth about culture, sovereignty, or civilizational boundaries. It was a scaling response. The regional model emerged because the original centralized arrangement could no longer handle global growth. What began as an operational convenience was later treated as if it were a permanent political architecture. That was the mistake. An administrative answer to scale was slowly transformed into a theory of legitimate authority.

That matters because once a scaling mechanism is reinterpreted as a political foundation, the registry layer begins to claim meanings it was never given. A service region is not a sovereign community. It is not a unified political people. It is not a coherent moral constituency. It is an administrative map. The fact that radically different states, legal systems, cultures, and even active geopolitical adversaries can sit inside the same RIR region should be enough to end the fantasy that these regions are natural political units. They are not. They are inherited service geographies. A colored registry map is not a theory of legitimate rule.

This also clarifies a point that current commentary often gets badly wrong. What is happening at AFRINIC is not evidence that “Africa” is inherently dysfunctional, and it is also not evidence that AFRINIC somehow embodies Africa in sovereign form. Both claims are false. AFRINIC is a registry institution within the RIR system. It is not Africa itself, and it does not speak as the legal owner of all interests within its service geography. Nor does applying for resources through AFRINIC amount to “taking Africa’s property.” AFRINIC is a distribution and registration channel inside a global system; it is not the original sovereign source of the assets it administers. The AFRINIC crisis is therefore not an African civilizational problem and also not proof of African political ownership being violated. It is a simply registry-structure problem: too much concentrated gatekeeping, too little accountability, too much mythology about representation, and too little respect for the networks that actually operate the resources and bear the downside.

Once that is understood, the asset question becomes much simpler than the rhetoric around it. Networks build businesses around number resources. They deploy capital, engineers, routers, contracts, support systems, customer relationships, financing structures, and operational discipline around them. They bear the downside when continuity is impaired. They bear the commercial damage when transferability is obstructed. They bear the operational consequences when recognition is destabilized. The registry does none of this. The registry records, coordinates, and services. It does not operate the network. It does not fund the network. It does not absorb the network’s downside. The only economically coherent presumption is therefore simple: the network that lawfully holds the resource, operates it in production, finances it, and bears the risk of it is the real asset holder in substance. Coordination is a service layer, not a superior ownership claim. Record-keeping is not sovereignty. Administration is not title.

This is exactly why the language of “stewardship” becomes dangerous when left imprecise. A narrow coordination function can be legitimate. A bookkeeping and registration function can be useful. But once coordination is inflated into stewardship, and stewardship is inflated further into political entitlement, the registry function mutates into something it was never meant to be. A registry is not a state. A continent is not a sovereign legal person. An RIR is not authorized to convert operator-held assets into political property through rhetoric about “the region,” “the people,” or “public destiny.” The moment that move is accepted, the numbering layer begins to attack the architecture of the Internet from below. It replaces operational continuity with political exposure.

That exposure is no longer hypothetical. A fragmented coordination layer already requires cross-regional choreography, separate policies, and synchronized administrative action. That may be tolerable as bureaucratic overhead. It is not tolerable as the basis for a superior political claim over globally used assets. A fragmented coordination layer is one thing. A fragmented coordination layer claiming higher moral or political title over the assets used by networks is something else entirely. The first is administration. The second is overreach.

The accountability problem becomes even sharper when placed next to the contract terms. My earlier argument on this point remains decisive. Registry institutions with highly limited downside are positioned above resources with very large real-world economic consequences. That was already unstable when framed as technical administration. It becomes intellectually indefensible when the same institutions, or the insiders around them, begin to imply some wider regional or public-trust claim over the underlying assets. That is no longer just power detached from accountability. It is power detached from accountability while pretending to rest on a civilizational mandate.

This is why the current insider rhetoric should be resisted at its root. An RIR can coordinate resources for its members. That is a narrow and intelligible function. It is not the same as representing all states, all peoples, all political interests, or all moral claims inside a service geography. The distinction is fundamental. The moment representation of members is rhetorically inflated into representation of a region itself, the system begins to drift into fiction. No serious person can explain how a single registry layer can politically “represent” states with radically different interests, legal systems, and geopolitical conflicts simply because they appear within the same inherited service map.

This is also why my earlier AFRINIC-election argument about clique control and decentralisation remains relevant. The problem was straightforward: too much decision-making power concentrated in a small circle, too little transparency in how that power was exercised, and too much ability for insiders to use process, procedure, and institutional familiarity to control outcomes against the broader membership. The answer was not more rhetoric. The answer was decentralised, member-led governance and stronger recognition of address holders’ rights. That was not campaign language. It was a structural argument. Once scarce and economically meaningful coordination power is concentrated in a small procedural class, complexity becomes a weapon, process becomes a filter, and “community” becomes a word used mainly by those already closest to the microphone.

But decentralisation does not mean replacing registry overclaim with state overclaim. Internet number resources are not national trophies. They are not continental entitlements. They are not raw material for political theatre. The Internet became humanity’s first globally scaled decentralized communication network precisely because no single state or alliance could simply define the whole. To reinsert state-style ownership logic into the numbering layer is to undermine that achievement. If a government, a coalition of governments, or a registry-adjacent political class claims that the addresses operated by networks in some region are really “the people’s property” to be politically administered from above, that is not liberation. It is expropriatory logic wearing more fashionable language.

Nor is the answer a single global monopoly. If five regional monopolies can drift toward overclaim, mythology, and politicization, one planetary monopoly would merely scale the same failure mode to maximum size. The problem is not just that the map is wrong. The deeper problem is concentrated discretionary power over globally relevant and economically meaningful resources. Splitting that power into regional silos did not solve it. Centralizing it further would solve even less.

The only serious destination is decentralisation with interoperability: globally compatible standards, portable recognition, auditable records, transparent rules, and continuity guarantees strong enough that no single registry institution, region, state bloc, or insider circle can place large amounts of economic life at risk by reinterpretation or capture. The design target should be verifiability, portability, and continuity, not mythology about geography. The real question is not which political story feels most emotionally satisfying. The real question is which structure keeps the Internet stable while minimizing arbitrary power over the resources on which networks actually depend.

The practical test is simple. If a doctrine says that the network which built, paid for, operated, and bore the downside of a number resource does not really hold it, but some regional administrative layer does, that doctrine is false. If a doctrine says that a registry and its immediate procedural circle can speak for all the peoples inside a service region that spans radically different states and even active conflicts, that doctrine is false. If a doctrine says that applying through an RIR service channel is equivalent to taking a continent’s property, that doctrine is false. And if a doctrine says that the failure of one registry institution should be attributed either to the defects of a civilization or to the wounded sovereignty of a region, that doctrine is false.

For most of the Internet’s earlier life, the regional registry model could be tolerated as a useful scaling compromise. That is no longer enough. Once number resources became scarce, transferable, financially meaningful, and embedded in real operating businesses, the standard changed. The registry layer must now justify itself under the disciplines of accountability, portability, and continuity. It cannot substitute geography for legitimacy. It cannot substitute process for ownership. And it cannot transform operator-held assets into political property by rhetorical inflation.

Internet number resources are not political property. They are operator-held assets embedded in functioning networks. The registry may record them, coordinate around them, and help keep the system legible. But it does not become their sovereign, their moral superior, or their true owner. The moment the Internet accepts that fiction, it begins to dismantle the logic that made the network resilient in the first place.