Summary

  • The 2003 Geneva Declaration recognized the sovereign right of states over Internet-related public policy. It did not vest states with title to IP address space, appoint national number registries or convert global registrations into territorial concessions.
  • Public law normally reaches a person, company, licence, service, transaction, facility or harmful act through a defined jurisdiction. A number record performs a different function: it helps independent networks avoid conflicting use of globally unique identifiers.
  • Geography is relevant evidence but an unreliable ownership rule. Address blocks can support multinational networks, anycast services, roaming users, foreign customers and infrastructure that changes location without changing the registered resource.
  • Regulators can lawfully require domestic operators to maintain accurate information, manage security risk, respect rights, preserve continuity and comply with reasoned orders. The order must identify whom it binds, the legal basis, the review path and the operational act required.
  • A durable settlement separates regulation of conduct from custody of the registration asset. It preserves a globally legible uniqueness record while exposing registries and operators to applicable law, due process, correction, portability and proportionate remedies.

The sovereign claim begins with a valid power

The category error is difficult to correct because it begins with a proposition that is true. Governments do possess authority that private standards groups, registry associations and network operators do not. A legislature can create binding duties. A regulator can issue a licence and attach conditions authorized by law. A court can compel a party, set aside an unlawful decision and award a remedy. An enforcement body can investigate conduct using powers that voluntary associations do not have.

Internet infrastructure is not exempt from those powers merely because packets cross borders. A domestic carrier can be required to meet resilience standards. A company can be ordered to preserve evidence, protect personal data, disclose a security incident or stop an anticompetitive practice. An address broker can be subject to tax, fraud and sanctions law. A public authority can procure connectivity on conditions that influence routing security and continuity.

The mistake occurs one step later. Because public law can regulate conduct involving Internet numbers, the numbers themselves are described as sovereign property, national resources or territorial possessions. The government then appears to acquire a superior claim to the shared registration record. That inference does not follow. Jurisdiction over a regulated actor is not ownership of every coordinate the actor uses.

A state can regulate an airline without owning the global aircraft registration system, regulate a bank without owning the international payment message space, and regulate a mobile provider without owning every protocol identifier used by its equipment.

Public authority is strongest when its entity is precise. It becomes less credible when a lawful power over domestic conduct is expanded into an undefined claim over a global coordination facility.

Geneva allocated a public-policy role, not an address pool

The 2003 Geneva Declaration of Principles is the text most likely to support an inflated reading. Paragraph 49 says that policy authority for Internet-related public policy issues is the sovereign right of states. It also recognizes continuing roles for the private sector, civil society, intergovernmental organizations and international organizations in technical standards and related policy.

The sentence matters. It rejected the idea that technical institutions could define the whole field and leave governments outside. States did not surrender their constitutional, legislative or international responsibilities when Internet services became global. Public-policy questions remained public-policy questions.

But the Declaration did not assign states every function touching an Internet issue. It did not say that governments owned the DNS root, IP address pools, autonomous system numbers or routing tables. It did not create national credentials for the IANA number registries. It did not specify a transfer procedure, a hierarchy of sovereign claims or a rule for deciding which country owned a block used in several places.

That silence was not a drafting accident. Paragraph 49 distributed roles rather than collapsing them. Paragraph 50 asked for further study of Internet governance because the relationship among technical and public-policy functions was unresolved. The 2005 Tunis Agenda repeated both the sovereign-right sentence and the formulation that actors operate in their respective roles.

The sound reading is therefore bounded: states retain public-policy authority where law gives them jurisdiction and responsibility. The text does not convert every technical function into an entity of sovereign ownership. The boundary is part of the settlement, not an exception to it.

The regulated entity and the registration asset are different things

Administrative law begins by identifying what a decision is about. A regulator may be controlling a licensed service, the conduct of a corporate person, use of a facility, compliance with a security duty, a market transaction or harm to a protected interest. The legal instrument should name that entity and connect it to a statutory power.

The number registry has another entity. RFC 7020 describes a system for distributing globally unique IP address space and autonomous system numbers and maintaining accurate registration information. Uniqueness means that the same number is not allocated to more than one party at the same time. The record helps operators know which claim the established coordination system recognizes.

That record can be economically and operationally important. It can affect transfers, reverse DNS, contact data, route-security credentials and confidence in a network's claim. Calling it merely clerical would understate the consequences of changing it. Yet importance does not alter its category. It remains a coordination asset: a shared record intended to be intelligible across institutions and borders.

The legal entity and registration asset can interact. A court may decide that a company validly acquired rights under a transaction. A regulator may find that a licence holder obtained addresses by fraud. A registry may need to update its record to reflect a binding decision. The interaction requires a documented handoff. It does not make the registry the court, or the court the global registry.

Once this distinction is maintained, government power can be respected without inventing national ownership of the uniqueness ledger.

A unique identifier is not a slice of territory

Territorial reasoning is attractive because addresses sound geographic. Postal addresses point to places. Telephone country codes have national associations. Regional registries publish service maps. Network operators describe markets by country. The language encourages a picture in which IP numbers lie inside borders.

The running network does not follow that picture. A company incorporated in one state may receive resources through an organization serving a much larger region. It may announce the same prefix from facilities in several countries. Anycast can make one address reachable through multiple locations. A cloud provider can move workloads while retaining addresses. A carrier can serve roaming users, foreign subsidiaries and international customers. An enterprise can hold a block even when no route is currently announced.

The registry itself is hierarchical, but the hierarchy is not territorial title. IANA manages the top of the current allocation hierarchy; Regional Internet Registries serve broad administrative regions; local registries and other customers serve networks and users. The hierarchy helps distribute a finite identifier space while retaining a coherent record. It does not divide the Internet into five sovereign estates.

Geography remains relevant. It can determine jurisdiction, service eligibility, consumer exposure, licence requirements, tax obligations and which registry relationship is practical. It can help identify abuse and assess whether a claim is credible. The error is to turn relevance into ownership.

A border can tell a regulator which company it may lawfully command. It cannot, by itself, tell every network in the world which claim to an IP number it must recognize.

Four acts are routinely compressed into one word

Claims about control become clearer when four different acts are separated. The first is legal regulation: a public authority creates or applies a duty to a person subject to its jurisdiction. The second is adjudication: a court or competent body determines rights or liabilities between parties. The third is registration: the recognized recordkeeper changes the information associated with a number resource. The fourth is routing: autonomous networks announce, propagate, filter and prefer reachability information.

These acts may point in the same direction, but they are not interchangeable. A regulator can order a domestic operator to stop announcing a prefix. That order may be effective against the operator even before a registry record changes. A court can determine that a transaction is valid, while a registry still needs authenticated documents and a coordinated cutover. A registry can update a holder record, while networks independently decide whether to accept the resulting route. A network can announce a prefix without a lawful claim, proving that operation alone is not title.

Institutional language often hides these distinctions. A government says it has “control” of national Internet resources. A registry says it “manages” the address space. An operator says it “owns” a block. Each verb may refer to a different bundle of acts and expectations.

The remedy is not to ban ordinary words. It is to demand an executable sentence. Who must do what, under which instrument, to which record or system, by when, with what review and what consequence? If the sentence cannot be completed, the sovereignty claim is performing political work without identifying an operative power.

Public law reaches legal persons before it reaches global records

A national authority ordinarily acts through a jurisdictional connection. The person is incorporated, licensed, resident, established or doing business in the state. The conduct occurs there, causes a recognized effect there or falls within a carefully defined extraterritorial rule. The affected party receives whatever procedure the legal system requires.

This person-centred route is not a weakness. It is what makes coercion reviewable. A carrier can ask whether the regulator had statutory power. A court can examine evidence, proportionality, reasons and consistency. Other affected parties can seek participation or relief. The duty has an addressee.

By contrast, a declaration that “all IP addresses used in the country are national resources” has no obvious stopping point. Does “used” mean announced from a domestic router, assigned to a customer, visible to users, hosted in a data centre, registered to a local company or geolocated by a commercial database? What happens when several definitions apply to different states? Does the declaration bind a foreign RIR, a multinational holder, a route collector or every autonomous network?

The broader formulation feels more sovereign but is less administrable. It abandons the legal person and substitutes a movable technical coordinate. The result is likely to be selective enforcement against actors within reach while the universal ownership claim remains symbolic.

Good public law should state the narrower truth. The state regulates specified actors and conduct under law. It may require those actors to take steps concerning number resources. The global effect depends on the lawful order reaching the institutions and systems capable of implementing it.

The legitimate regulatory toolkit is already powerful

Rejecting state ownership of the global number record does not require a deregulated registry layer. Governments have a substantial toolkit that can be used without territorializing the ledger.

A telecommunications regulator can attach conditions to licences concerning continuity, emergency contact, accurate network information and security controls. A competition authority can examine foreclosure, discriminatory access, collusion and abuse of a bottleneck. A data-protection authority can regulate publication and handling of personal registration data. Courts can decide contract, insolvency, fraud and property-related disputes between parties. Legislatures can establish due-process duties and remedies for high-consequence infrastructure decisions.

Public procurement offers another lever. Governments buy connectivity, cloud services and managed networks. They can require suppliers to document resource control, maintain route-origin authorizations, plan for registry failure and preserve service through a dispute. Financial supervision can require companies to disclose concentrated dependency on a registry or a leased address portfolio. Consumer law can address misleading claims about continuity and ownership.

These powers target identifiable risks. They can be tested against evidence and corrected through ordinary institutions. They also leave room for technical diversity: a state can specify the resilience outcome without pretending to administer every global allocation.

The political attraction of registry ownership is that one sweeping claim seems to replace many difficult regulatory tasks. In practice, it replaces enforceable duties with a slogan. A regulator serves the public better by using its real powers precisely than by claiming a global asset it cannot maintain alone.

Property language contains several questions, not one answer

IPv4 scarcity gives the ownership debate economic urgency. Address blocks are transferred, leased, valued, financed and relied upon in transactions. Networks spend money to acquire control and build services around them. Registry decisions can therefore affect assets with consequences far beyond an annual service fee.

It does not follow that one universal property category has already settled every issue. A transaction can include contractual rights, registry recognition, credentials, operational control, customer commitments and expectations protected by domestic law. Different jurisdictions may characterize parts of that bundle differently. A holder's strong economic interest does not prove that a state owns the resource. A registry disclaimer does not make the holder's reliance disappear.

The practical distinction is between the underlying interests and the registration asset that coordinates them. A court can protect a creditor or buyer. A regulator can prevent fraud. A registry can preserve one recognized state and record a dispute. Networks can decide what to route. Each institution should state which part of the bundle it is deciding.

Political property claims are especially dangerous because they skip this analysis. “National resource” can imply ownership by the state, stewardship for the public, regulatory jurisdiction, strategic importance or simply a desire that domestic users receive fair access. Those propositions have different legal consequences.

The phrase should never do the work of a statute, judgment, contract and registry update at once. If public ownership is truly intended, the law must define the property, acquisition, compensation, affected parties, cross-border effect and review. Without that detail, the label adds conflict without adding authority.

National slices would destroy the function they claim to protect

The uniqueness record is useful because entities can consult a shared chain of registration and avoid simultaneous conflicting assignments. If every state treats the numbers visible within its borders as its own political property, the same address can attract several national claims.

Consider a multinational content provider announcing one prefix through facilities in four countries. The company is incorporated in a fifth, holds its registry relationship in a sixth and serves users everywhere. A rule based on physical infrastructure points to four owners. A rule based on incorporation points to one. A rule based on users points to nearly every country. A rule based on the registry's location points to another. None is naturally superior for preserving global uniqueness.

The conflict cannot be solved by saying each state is sovereign inside its borders. Routing a globally unique address is not confined to the border. Other networks must decide which announcement to accept. A domestically valid reassignment can create a collision outside the state if the prior global record and route remain active.

Territorial duplication would also burden smaller states. Large markets could use regulatory and commercial pressure to make their preferred records more widely recognized. Countries with less routing influence would possess formal sovereignty but weaker practical effect. The result would not be equality; it would be competition among incompatible registers weighted by market power.

The public interest in one Internet therefore supports a thin common function. The ledger must remain globally legible enough to prevent conflicting ordinary use. Public authorities can regulate around that function, challenge decisions and require remedies. They should not partition the very property that makes the function valuable.

A service region is not a constitutional territory

Regional Internet Registries complicate the analysis because their names and maps look political. AFRINIC, APNIC, ARIN, LACNIC and the RIPE NCC serve large geographic areas. Their communities develop policies, select officeholders through different arrangements and describe regional needs. Governments and companies often attend their meetings.

The geography has administrative value. It distributes work, creates more accessible policy venues and permits rules to respond to different operating conditions. RFC 7020 describes the RIRs as operating in continent-sized geopolitical regions. The Number Resource Organization describes five separately constituted organizations and their service areas.

Neither description turns a service region into a sovereign people. The countries within one RIR area have different constitutions, regulators, languages, economic interests and foreign policies. A registry corporation is created under the law of a particular jurisdiction, not by a regional constitution ratified by every resident. Its entities do not form a continental electorate.

This boundary cuts both ways. A registry should not claim political ownership on behalf of “the region.” A government should not claim the registry because it happens to be incorporated or hosted locally. Members can authorize specified institutional acts under bylaws and agreements. States can exercise public power under law. Neither authorization converts the administrative map into title over the people, networks or number resources within it.

Regional coordination is legitimate to the extent that it performs a bounded function accountably. Geography helps organize the service. It does not supply a missing sovereign foundation.

The host state has jurisdiction, not beneficial ownership

Every registry organization exists somewhere in law. It has a corporate form, bank accounts, employees, contracts and records. Courts in the place of incorporation or operation can issue orders. Regulators may apply employment, tax, data, insolvency and corporate law. This is real public authority over a critical institution.

The host state's position creates both necessity and risk. Someone must be able to supervise the legal person, enforce judgments and resolve corporate failure. Yet orders affecting a registry can have consequences for networks and holders in many other countries. A domestic dispute can therefore become a cross-border continuity problem.

The correct conclusion is not that the registry floats above law. It is that its globally relied-upon functions need continuity arrangements proportionate to the host-state exposure. Records should be exportable and independently verifiable. Credentials and service dependencies should not sit behind one unrehearsed point of failure. A receiver or court should be able to preserve the registry function without inheriting a claim to rule the service region.

Beneficial ownership does not arise from hosting. The state does not become owner of every registered block because its courts can bind the corporation maintaining the database. Equally, the registry cannot invoke global importance to evade a valid local judgment.

The institutional design task is to hold both propositions at once: ordinary law applies to the organization, and the coordinated record must be capable of surviving a local legal crisis. That is a continuity duty, not a sovereignty transfer.

A conflict between two lawful orders exposes the missing rule

Imagine a company incorporated in State A, licensed to operate a network in State B and using address space registered through an organization incorporated in State C. A court in A approves a restructuring that transfers the operating assets. A regulator in B orders continuity for local customers and prohibits an abrupt routing change. A court in C freezes changes to disputed registry records while ownership litigation proceeds.

Each order may be lawful within its own field. None alone answers every question. The restructuring order determines rights among parties before that court. The continuity order governs the licensed service in B. The freeze binds the registry in C. Network operators elsewhere retain their own route policies.

Calling the addresses sovereign property of A, B or C does not resolve the conflict. It merely chooses a winner without a choice-of-law analysis, recognition process or technical cutover plan. The useful response is functional coordination.

The parties need an authenticated record of the competing orders, a temporary continuity arrangement, a defined custodian for credentials, notice to affected networks and a forum capable of deciding the disputed legal interest. The registry should identify what it can preserve and which binding instruction it requires. The regulator should distinguish continuity of domestic service from permanent title. The court should understand which technical acts its order reaches.

This example shows why the global ledger cannot be treated as a political prize. Its role during conflict is to preserve a coherent, reviewable state while competent institutions decide their parts of the dispute. Territorial ownership would make simultaneous obedience impossible.

Registries are subject to law without becoming public authorities

Private coordination can exercise public consequences. A registry may deny an application, suspend a service, refuse a transfer or alter security-related state. Those acts can affect connectivity, competition and valuable reliance. Private legal form should not insulate them from applicable law or serious procedural standards.

But exposure to public law does not automatically transform the registry into a government department. Its authority may arise from contracts, membership rules, recognized practices, technical dependency and acceptance by networks. The specific basis matters because it determines the available remedy.

If a registry breaches a contract, the affected party needs contractual and judicial recourse. If its decision produces anticompetitive effects, competition law may apply. If it mishandles personal data, privacy law may apply. If the corporate body fails, insolvency and corporate law apply. If a policy process excludes affected interests, institutional review and public criticism may be appropriate.

The ICANN Bylaws offer a useful limiting statement at the coordinating level. They define ICANN's mission concerning unique identifiers, including top-level number coordination, while stating that ICANN does not hold governmentally authorized regulatory authority. The limitation does not make ICANN unaccountable. It identifies the kind of authority it does not possess.

RIRs need equally legible boundaries. They should be able to explain which act rests on policy, which on agreement, which on law, which is technical necessity and which remains discretionary. That explanation lets public law supervise private power without pretending the private body is sovereign.

Due process must follow the implementation chain

An adverse decision concerning number resources may move through several institutions before its full effect appears. A regulator issues an order. A licence holder changes a route. A registry updates a record. A certification system changes an authorization entity. Other networks alter validation or filtering. If procedure exists only at the first step, later implementation can create harm no one reviews.

A serious process should identify the affected resource, the evidence, the legal basis, the decision maker and every material implementation step. Notice must reach the holder and other parties whose rights are directly at risk, subject to a narrow emergency exception. Reasons should distinguish proven facts from technical prediction and policy judgment.

Review must be timely enough to matter. An appeal decided after credentials are revoked, records propagated and customers disconnected may be formally available but practically empty. A stay should be possible where preservation will not create a demonstrable security or uniqueness risk. Irreversible steps require a higher explanation.

The registry receiving a government order should publish or securely preserve enough information to establish provenance, scope and status. It should not translate a direction to one domestic company into a global ownership claim unless the order actually requires and lawfully supports that result. It should record conflicts rather than silently erase them.

These are ordinary administrative-law instincts applied to a distributed implementation chain: authority, notice, reasons, proportionality, review and effective remedy. They make sovereignty operationally credible by limiting it to acts the state can justify.

Registration is strong evidence but incomplete title

A current registry entry can be powerful evidence. It shows what the established coordination system recognizes, often after authentication and policy review. Historical records can show continuity, transfers, contacts and prior decisions. Security services may rely on the registered relationship.

Yet the entry does not answer every legal and operational question. It may be stale, disputed or based on information later shown to be false. A holder can authorize another network to originate a route. A company can acquire contractual interests before all records change. A court can bind parties in ways the registry has not yet implemented. A route can exist without valid authority.

Treating registration as conclusive state title creates one error; treating it as meaningless clerical data creates another. The appropriate status is rebuttable, function-specific evidence with a defined correction path.

Public authorities should say what they infer from the record. It may identify the recognized holder for registry services, support control verification or establish who received notice. It does not necessarily prove beneficial ownership under every law. Registries should say what their entry attests and what it does not.

This bounded evidentiary approach also protects uniqueness. A court can consider other evidence without immediately authorizing a second conflicting registration. A dispute marker, stay or controlled transition can preserve one effective state while legal rights are resolved.

The registration asset is therefore neither sovereign deed nor disposable note. It is a consequential shared record whose evidentiary force depends on accuracy, procedure and acceptance.

Public-interest language needs an identified beneficiary and remedy

Claims of national ownership are often framed as protection of the public. Scarce addresses should benefit domestic development. Critical networks should not depend on foreign institutions. Smaller operators should have fair access. Government should be able to protect connectivity in a crisis. These are legitimate concerns.

The phrase “public interest” does not decide how to meet them. A policy reserving resources for domestic applicants may help entrants or may entrench licensed incumbents. A national register may improve visibility or create another political gatekeeper. A transfer restriction may preserve local capacity or strand an asset needed by a cross-border network.

Institutional design should identify the beneficiary, mechanism, cost and remedy. If the goal is continuity, require tested failover and portable records. If the goal is competition, address discriminatory allocation and market entry. If the goal is national security, define the threat and impose proportionate duties on critical operators. If the goal is development, measure whether smaller networks actually receive usable capacity.

The state should also disclose conflicts. It may be regulator, network owner, major customer and security authority at the same time. A public enterprise seeking addresses should not receive an unreviewable advantage because its owner also defines the national interest.

Public-interest legitimacy comes from lawful objectives, evidence, fair procedure and effective remedy. Ownership language is not a shortcut. It can conceal who gains and who loses behind the prestige of sovereignty.

The ledger needs public-law interfaces, not a political owner

The alternative to national ownership is not institutional immunity. A globally legible registry should have designed interfaces with public authority.

There should be a clear channel for authenticated court and regulatory orders, including scope, legal status, duration and contact for review. The record should distinguish an uncontested registration from one subject to a preserved claim. Emergency action should be logged and revisited. Holders should receive notice and reasons unless a lawful temporary restriction applies.

Registry operators should publish the legal entities, jurisdictions, agreements and continuity arrangements through which they act. They should explain how orders from the host state are handled and how conflicting foreign judgments are assessed. Critical records should be exportable in a form that permits independent integrity checking and service restoration.

Public authorities, in turn, should direct orders to actors and acts they can lawfully reach. If they seek a registry change, they should identify the recognition or enforcement route and account for third-party effects. Courts should have technical evidence before requiring irreversible changes.

No single political owner is required for these interfaces. Indeed, an owner with final discretionary power would create the very concentration that cross-border coordination is meant to avoid. The objective is a record capable of receiving lawful decisions without becoming the property of the institution that transmits them.

This arrangement is less dramatic than sovereignty rhetoric. It is also more enforceable. It gives each institution a duty it can perform and a failure that can be reviewed.

Operator rights are the missing bridge

The government-versus-registry debate often leaves the operator as an entity rather than a rights-bearing party. Yet the operator deploys the addresses, serves customers, bears outage costs and supplies much of the evidence needed to assess control.

A credible framework should give the recognized holder access to its complete record, notice of material adverse action, reasons, a correction process, timely review and a practical stay before irreversible change. It should permit an authenticated export of registration history and credentials needed for continuity. Liability should be proportionate to the foreseeable consequence of wrongful action.

These rights limit both private and public overreach. A registry cannot hide behind community language when it changes a high-consequence record. A regulator cannot hide behind sovereignty when it directs an opaque seizure. Each must identify its power and allow challenge.

Rights also improve evidence. Operators have incentives to keep records accurate when correction is usable and continuity depends on them. They are more likely to report disputes early when doing so does not trigger arbitrary loss. Review bodies receive a fuller account than a contest between institutional assertions.

Portability deserves particular attention. It should not create duplicate authoritative state or permit evasion of a valid order. Properly designed, it allows the service relationship and verifiable record to survive institutional failure or abuse while carrying existing disputes forward. The possibility of lawful substitution disciplines the recordkeeper without pretending that every holder may invent its own truth.

The bridge between sovereign law and global coordination is not ownership by either side. It is enforceable rights at the point where decisions affect the network.

Sovereignty should mean responsibility for consequences

Political debate often treats sovereignty as freedom from outside constraint. Administrative law offers a more demanding account. Public power carries duties: legality, reasons, equality, proportionality, review and responsibility for foreseeable consequences.

Applied to Internet numbers, this means a government should assess the cross-border effects of an order, not merely assert jurisdiction. It should distinguish urgent protection from permanent disposition. It should provide a route for affected parties to contest mistakes. It should coordinate where unilateral action would create conflicting registration or widespread outage.

The same discipline should apply when a state participates through a regional or global forum. A diplomatic statement may set policy. An advisory position may influence a private coordinator. Neither should be cited later as if it were a court judgment or technical implementation. Government representatives should identify which authority they are exercising in each setting.

Responsibility also includes institutional capacity. If a state seeks direct custody of critical records, it must show how it will maintain integrity, continuity, security, neutrality and cross-border recognition. Political entitlement without operational capacity increases the risk to its own citizens.

The mature sovereign position is therefore not “the numbers are ours.” It is “we will use lawful powers to protect people and infrastructure, and we will cooperate to preserve the shared registration function on which those protections depend.” That claim is narrower in rhetoric and stronger in law.

Watchpoints for the next sovereignty claim

The first watchpoint is vocabulary. “National resource,” “digital sovereignty,” “strategic asset” and “public ownership” should trigger a request for the exact legal entity. Is the claim about a licensed service, a holder's rights, a registry database, an unallocated pool, routing behavior or state procurement?

The second is the implementing actor. A ministry statement is not a registry update. A registry update is not a route. A diplomatic resolution is not domestic law. The party controlling the relevant act should be named.

The third is collision risk. Any proposal for national custody should explain how it prevents two states from authorizing conflicting use of the same number, especially for multinational and anycast networks.

The fourth is procedure. Look for notice, reasons, evidence, a stay, review and correction before an adverse change. Emergency clauses should expire or require renewed justification.

The fifth is institutional self-interest. A state-owned carrier, incumbent operator, registry organization or political coalition may gain from a sovereignty claim. The public-interest rationale should be tested against that distribution.

The sixth is continuity. A proposal should preserve verifiable history, credentials, dispute status and service if the responsible institution fails. Replicating political control without a tested recovery path is not resilience.

The seventh is reciprocity. A state should ask whether it would accept the same doctrine when another government claims numbers used by its own companies abroad. A rule that works only when one's own jurisdiction is the claimant is not a global coordination rule.

Conclusion: regulate the conduct, preserve the coordinate

Government sovereignty and global number coordination are not enemies. They answer different institutional questions.

The state asks who may lawfully provide a service, what duties protect the public, how markets remain fair, what conduct is prohibited and which remedy follows a breach. Its authority can be coercive because it rests on a legal order with institutions for legislation, administration and review.

The number registry asks which claim to a globally unique identifier is recognized in the shared record, how that claim changes, how conflicts are marked and how independent networks obtain accurate information. Its authority depends on bounded function, reliable operation, accepted procedures and the willingness of others to rely on the result.

When the state regulates an operator, the number may be part of the facts. When a court decides a transaction, the registry may need to respond. When a registry makes an adverse decision, public law may constrain it. Interaction is inevitable. Ownership merger is not.

Territorializing the record would make sovereignty less effective, not more. It would create incompatible national claims, reward the states with the greatest routing and market power, and leave smaller countries with symbolic title to a fragmented system. It would also obscure real remedies by treating every problem as a struggle for custody.

The better settlement is demanding but intelligible. Governments use real law against actors and harms within lawful reach. Registries maintain one legible state, expose their power to correction and preserve continuity. Operators receive enforceable rights and bear duties proportionate to their control. Courts identify the parties and effects their orders bind. Networks retain responsibility for routing decisions.

Sovereignty should protect people from unaccountable power. It should not become a reason to invent another unaccountable owner. Regulate the domestic conduct. Adjudicate the legal dispute. Require the technical act. Preserve the globally unique coordinate. The legitimacy of each institution depends on knowing which of those tasks it is actually performing.

Sources