Summary

  • The five regional Internet registries administer one coordinated class of global resources, yet their published dispute routes differ in scope, reviewer, timing, confidentiality, cost allocation, court access and protection of the status quo.
  • Geography is not an ordinary forum choice. Most holders cannot move a disputed registration to another registry merely to obtain a cheaper appeal, a more independent reviewer or a faster continuity order.
  • Internal escalation, membership oversight, ombuds-style assistance, arbitration and court proceedings perform different jobs. Treating any one of them as a complete accountability system leaves predictable gaps.
  • The decisive remedy is often temporary: reasons, record access and a narrowly framed pause can preserve registration services while the merits are tested, whereas a successful ruling after irreversible reissue may have little practical value.
  • A common baseline should require notice, reasons, an independent merits review, affordable access, a continuity test, a complete decisional record and a route to urgent external relief, while leaving each region free to choose its law and institutional form.

One resource system, five remedial constitutions

An IP address block does not become technically North American when it is registered by ARIN, European when it is registered by the RIPE NCC, or African when it is registered by AFRINIC. Networks announce addresses across borders, customers reach them from every region, and counterparties evaluate their registration and routing status globally. The administrative relationship is regional; the operational consequences are not.

That distinction creates a constitutional oddity. The Internet Numbers Registry System coordinates uniqueness, conservation, aggregation and accurate registration through a hierarchy with five regional registries. Yet when a registration decision is contested, the route to relief changes with the registry. One institution begins with layered staff escalation and a defined request appeal. Another points members toward management review and, for broader disputes, commercial arbitration. A third uses a specialist panel of community arbiters.

Elsewhere, a service agreement directs a party toward a board appeal or a national conciliation and arbitration centre.

These are not cosmetic variations. They affect who may complain, which acts are reviewable, whether the reviewer is institutionally separate, who sees the record, how quickly a claim must be filed, who pays, whether reasons become public and what happens to services before a decision. Two operators presenting similar evidence can therefore face materially different remedies.

The resulting inequality is usually accidental. An operator chose a region because that is where it qualified for service, not because it preferred one law of interim relief over another. Registration geography has nevertheless become remedial geography. A globally coordinated system should be able to tolerate regional legal diversity without making practical due process a regional lottery.

Remedy shopping without a shop

The phrase "remedy shopping" normally suggests strategic choice. Litigants select a forum whose law, procedure or judges appear favourable. Number-resource holders often possess no comparable menu. Regional service criteria, organisational location, contractual history and legacy arrangements largely determine the registry relationship before any dispute exists. Once the contested act occurs, moving the registration is generally not an available way to change the reviewer.

The shopping happens instead through institutional design. Lawyers and operators compare internal appeals, arbitration clauses, membership votes, complaints channels and courts because they need to identify the route that can deliver the required result. They may discover that the internal route reviews only a denied request while the contract route covers suspension; that a community panel can recommend action but a court can compel it; or that emergency relief is available externally even though ordinary claims must first pass through arbitration.

This is shopping among remedies, not among registries. It consumes time precisely when time is expensive. A resource holder must determine whether to escalate to management, notify a corporate officer, invoke a service agreement, seek an interim order, preserve a limitation period or attempt settlement. A mistaken choice can expire a deadline or allow the operative registration status to change.

The registry has a repeat-player advantage. It knows which channel it regards as valid and which remedies it considers available. The holder may encounter the architecture once. A minimum baseline would not eliminate the need for advice, but it would ensure that every adverse decision arrives with a usable map and that no region omits the basic bridge between an internal objection and effective independent relief.

The technical commonality is narrower, and stronger, than a claim of ownership

The case for a baseline does not depend on describing Internet numbers as ordinary property. The five registries use contracts and policies with different formulations of a holder's interest, and national law may classify that interest differently. A due-process standard need not resolve those questions.

The stronger common ground is functional. RFC 7020 describes a coordinated hierarchy intended to maintain unique allocations and accurate registration information while serving operational needs. It also recognises that registry goals can conflict with the interests of individual users and resource consumers, requiring judgment and cooperation through community-developed policy. A dispute is therefore not surprising evidence of system failure. It is a foreseeable consequence of administration in a scarce, technically interdependent environment.

The need for remedies follows from that function. A decision about a request, transfer, registration record, contractual status, reverse delegation or related service can alter the evidence on which networks and counterparties rely. The registry does not control every route announcement, but its records and services can influence transactions, security operations and confidence. Correcting an error after those effects have compounded may be difficult.

A common baseline would attach to the exercise of registry power, not to a universal theory of title. It would ask whether an affected person received notice, reasons, the material record, independent review and a meaningful chance to preserve continuity. Those protections remain sensible whether the underlying interest is described as contractual, administrative, licensed, custodial or something else under local law.

ICP-2 supplies principles, not a complete appeal code

ICP-2, accepted in 2001 as criteria for recognising new regional registries, is an important starting point because it explains why regional authority is legitimate. It calls for open and transparent policy development, fair representation, neutrality and impartial treatment. It expects technical capability, financial independence, proper records and auditability. These requirements help justify the delegation of a globally important registration function to regionally rooted institutions.

But the document is not a code of civil procedure. It does not specify the minimum contents of an adverse notice, the independence of a reviewer, a filing period, cost protection, disclosure duties or a temporary hold. It was written to evaluate the establishment of a registry, not to determine every remedy available to a member decades later.

That gap has become more consequential since 2000. IPv4 scarcity increased the commercial importance of transfers and registration status. RPKI and other registry-linked services added security consequences. Organisations operate across several regions. Sanctions, insolvency, corporate transactions and fraud controls create disputes that do not fit neatly into the older model of a disappointed applicant asking for more address space.

The correct use of ICP-2 is therefore constitutional rather than literal. Neutrality, auditability and reliable service imply that affected parties must be able to test a serious administrative decision. The exact tribunal can remain regional. The minimum qualities of access should not. A recognition framework that values impartial treatment at entry but says nothing about remedy after a contested act leaves half of accountability unstated.

The published map reveals real divergence

The Number Resource Organization's governance matrix is useful because it places the five institutions in comparable rows and links to their own governing materials. Its dispute-resolution row points to the AFRINIC service agreement, the APNIC by-laws, the ARIN appeal and registration agreement, the LACNIC registration agreement and the RIPE NCC conflict arbitration procedure. Its deregistration row points elsewhere again. The map itself demonstrates that no single instrument or vocabulary governs the field.

The matrix also carries an institutional description of the frameworks as robust and accountable. That claim is evidence of how the registries present their arrangements, not a conclusion that comparative analysis must adopt. A link to a procedure proves that a procedure is published. It does not prove that the procedure covers the decisive act, provides an affordable hearing or protects service while review is pending.

Comparison must therefore move below the existence test. For each registry, six questions matter. What triggers review? Who decides? What record is available? What can the reviewer order? Who pays? What happens before the ruling? A seventh question concerns the path outward: which court or arbitral body can preserve the status quo or enforce a result?

These questions reveal complementary strengths and gaps rather than a simple league table. A specialist panel may understand policy but lack coercive power. Binding arbitration may deliver finality but impose cost and confidentiality. A management escalation may correct mistakes quickly but remain internal. Courts may issue urgent orders but require local counsel and a justiciable cause. Good design combines functions instead of announcing one route as sufficient.

ARIN separates a request appeal from the wider contractual dispute

ARIN's published resource-request appeal is comparatively explicit about its initial scope. It is available to an organisation that believes staff did not follow community-established policies and procedures when reviewing a number-resource request. The administrative point of contact must first take the request through successive registration-services escalation, including senior operational levels. The written appeal then goes to the president and chief executive and general counsel within 30 business days, identifies the ticket, and proceeds under the registration agreement.

This route has a virtue: it tells an applicant how to challenge a denial and preserves the policies and procedures in force when the request was denied. It also exposes a limitation. A request appeal is not necessarily the same thing as review of suspension, termination, a database action, a service restriction or another contractual dispute. The affected party must read the wider agreement to understand the second constitution.

The current registration agreement provides cooperative settlement negotiations followed, after 30 days without success, by binding arbitration. Venue depends on whether the holder is principally in the United States, Canada or another country in the service region. Each side bears its own lawyers' fees, while the initiating party initially bears arbitral expenses. The agreement separately allows specified courts to issue temporary restraining orders, preliminary injunctions or other injunctive relief to maintain the status quo pending resolution.

This is a significant continuity bridge. It recognises that the merits forum and the emergency forum need not be the same. Yet access still depends on understanding two documents, choosing the correct claim and financing the initial step. A baseline should preserve ARIN's useful status-quo concept while making the relationship between request appeal, contractual arbitration and emergency relief unmistakable in every decision notice.

APNIC combines management escalation, member power and private arbitration

APNIC's public transparency material directs a person to a form for appealing an operational decision of the secretariat to APNIC management. That can be a sensible first correction mechanism. Managers can retrieve the file, identify a policy misapplication and reverse an error without the cost of formal proceedings. Its weakness is structural: management review remains review within the organisation that made the decision.

The by-laws supply wider machinery. They give members a high-threshold power to review or amend Executive Council decisions, a form of collective constitutional accountability. They also set out a dispute route covering matters arising under the by-laws, corporate instruments or agreements among specified institutional actors. Written notice is followed by nominated settlement representatives and a defined good-faith negotiation period. Unresolved disputes proceed to a single arbitrator under UNCITRAL rules. Information prepared and disclosed for arbitration is confidential.

Parties generally bear their own costs and share the arbitrator's charges unless the binding decision provides otherwise. Queensland law and courts form the legal setting, while interlocutory relief is excepted from the obligation to complete the preliminary route before legal proceedings.

This architecture performs several jobs, but they should not be confused. An individual operational appeal to management is not independent adjudication. A membership supermajority is not a practical remedy for a single holder facing an urgent service action. Confidential commercial arbitration can bind parties, but it produces less public learning unless outcomes are separately reported. Interlocutory court access may preserve continuity, but its availability and cost depend on the claim and forum.

APNIC illustrates why remedy labels are limited public evidence. The institution offers escalation, corporate democracy, arbitration and courts. The accountability question is whether an affected operator can move coherently among them before the disputed act becomes irreversible.

The RIPE NCC offers the clearest specialist middle layer

The RIPE NCC conflict arbitration procedure is the most visibly specialised of the five published routes. It uses a panel of arbiters expected to know the Internet environment, RIPE policies and RIPE NCC procedures. Board members and staff cannot serve where their functions prevent impartiality, and the general meeting has a role in approving the panel. The procedure covers defined disputes involving members, the RIPE NCC and some legacy-resource relationships, while excluding services whose terms already specify a competent national court.

Its name requires care. The procedure states that it is informal and is not arbitration in the Dutch civil-law sense. Parties can challenge a ruling in a competent national court. A request follows an unsuccessful effort to resolve the dispute directly and must be made within a reasonable period, no longer than one year. The procedure deals with information exchange, expert advice, timeframes, execution and a public case report.

Several design details are unusually concrete. If an outstanding invoice is disputed, payment can be held in escrow during the proceeding. The ruling can include implementation recommendations. Parties have two weeks to comply unless a party submits the dispute to a Dutch or other competent national court. Reasonable procedural costs are allocated to the losing party and capped below EUR 5,000 under the published rule.

These features make time, cost and transition more legible than a bare arbitration clause. They do not answer everything. Scope exclusions matter, external court proceedings remain costly, and a two-week transition can be demanding. Still, the specialist middle layer shows how regional expertise can narrow a case before ordinary litigation without pretending to replace national law.

LACNIC's agreement emphasises conciliation and final arbitration

LACNIC's published registration agreement places disputes in an Uruguayan legal and institutional setting. It directs the parties first to try conciliation through the conciliation and arbitration centre of the Uruguayan Chamber of Commerce. If conciliation fails, the controversy is to be resolved by arbitration under the chamber's rules, with three arbitrators.

The sequence values settlement before adjudication. That can reduce hostility and permit tailored outcomes: corrected records, revised deadlines, staged compliance or agreed technical steps. A three-person tribunal may provide deliberative depth for a complex dispute. But the agreement's concise clause leaves an ordinary holder needing further information about initiation, cost, language, appointment, urgent measures, publication and the treatment of services while conciliation is under way.

Conciliation also depends on bargaining conditions. It works best when both parties know that a timely independent decision will follow if talks fail. If the operator fears an imminent change and the institution controls the clock, negotiation can become a period in which leverage deteriorates. A continuity rule is therefore not hostile to settlement. It makes settlement more genuine by preventing irreversible action from deciding the case before the conciliator can.

The LACNIC example underlines a broader point. A contract can identify a final forum without providing a complete public route from the first adverse notice to that forum. The missing layer need not be another large tribunal. A clear notice, automatic record package, short internal reconsideration and published rule for temporary preservation could connect ordinary administration to the contractual mechanism while respecting Uruguayan arbitration law.

AFRINIC's board appeal exposes the meaning of finality

AFRINIC's registration service agreement contains a right-of-appeal clause framed around an organisation or member that believes the registry assigning its address did not perform properly. It requires written grounds to AFRINIC and the assigning registry within 21 days. The assigning registry is to forward relevant documents within 15 days, and the AFRINIC Board is to examine the appeal and issue a final decision within a further 15 days.

The timetable is attractive on paper. It promises a result in weeks rather than years and requires relevant documentation to move. Yet the wording raises questions for direct and indirect assignments, and the board remains part of the same corporate institution. "Final" can mean final within the contractual appeal, not necessarily immune from every remedy available under Mauritian law. A reader should not infer more than the instrument says.

The agreement's termination provisions add another layer. Before termination for breach, the applicant receives written notice and an invitation to show cause or cure, with 30 days to respond. That is a valuable pre-decision safeguard, but the agreement also describes immediate consequences once termination takes effect. The practical issue is whether an appeal, court application or other challenge pauses those consequences and under what conditions.

AFRINIC demonstrates the danger of equating speed with independence or finality with completeness. A quick board decision can correct ordinary mistakes. A conflict that implicates the institution's own authority, continuity or corporate capacity may require external supervision. The baseline should keep the quick route and add a clear bridge to independent review and narrowly tailored preservation.

Similar disputes are not identical disputes

Comparative work can become careless by flattening facts. A denied request for new IPv6 space is not the same as termination of a service agreement. A disputed transfer is not the same as a billing default. An RPKI service restriction is not identical to a change in public registration data. Different contracts, policies and national laws can justify different procedures and outcomes.

The claim here is narrower. When disputes engage comparable interests - an adverse registry decision, a contested application of policy, material operational consequences and a request for correction - the five systems expose affected parties to different procedural opportunities. The differences may determine whether the evidence is heard and continuity preserved before the merits are reached.

A baseline should therefore standardise questions, not outcomes. Every institution should identify the act, rule, evidence, decision-maker, effective time, review route and available temporary protection. Every affected party should have access to an independent merits assessment for a materially adverse decision. Every reviewer should be able to see the complete decisional record. Cost and timing should be proportionate to the risk.

Local law would continue to determine causes of action, arbitral enforcement and judicial remedies. Regional policy would continue to determine allocation criteria. The baseline would not tell a tribunal who wins. It would ensure that a comparable controversy reaches a competent decision-maker through a route that does not disappear because the registration sits on one side of a regional boundary.

Appeal, complaint, ombuds and arbitration are different instruments

Institutional documents often place several accountability channels on the same web page. That presentation can imply substitutability. In practice, each channel has a different constitutional role.

An appeal asks a person with authority to reconsider a decision and potentially replace it. A complaint may address service quality, delay, staff conduct or communication without reopening the substantive result. An ombuds function can help a person navigate the institution, investigate maladministration, mediate and identify systemic patterns; its strength is access and informality, while its usual weakness is limited coercive power. Arbitration produces a binding or contractually significant determination within an agreed jurisdiction.

A court can interpret law, compel evidence, issue injunctions and enforce or review awards, but at greater cost and formality.

Membership power is different again. Elections, meetings, petitions and votes can discipline boards and amend governance. They are essential for institutional legitimacy, yet they are poorly suited to deciding whether one transfer should pause on Friday afternoon. Collective oversight should improve the remedial constitution, not act as the emergency chamber.

A complete system links these tools. The complaint office identifies an error pattern. The ombuds or neutral officer helps frame a dispute and can recommend an immediate administrative hold. An independent reviewer decides the merits. Arbitration or court supplies enforceability. Members receive aggregate reporting and change the rules where repeated failures appear. Calling any single component "the appeal" obscures the gaps between them.

The pause often matters more than the eventual judgment

Registry disputes are unusually sensitive to sequence. A number block can be shown as revoked, a transfer can close, an account can lose service, a reverse delegation can change, or a certification service can be affected before an appeal concludes. The technical Internet may not stop at once, but counterparties react to registry signals. Banks, purchasers, transit providers, customers and security teams do not wait for a final judgment to update risk assessments.

A later victory may therefore be incomplete. If a block has been reissued, restoration can collide with uniqueness. If a transaction has failed, the buyer may have gone elsewhere. If customers have migrated, a corrected record does not reconstruct revenue. If an operator had to renumber or redesign security controls, legal success arrives after operational cost.

This makes temporary preservation a core remedy rather than an exceptional favour. ARIN's agreement expressly contemplates court relief to maintain the status quo. APNIC's by-laws preserve access to interlocutory proceedings. The RIPE NCC procedure uses escrow for a disputed invoice and creates a short period in which court referral affects execution. Other published clauses are less explicit about a universal pause.

The baseline should require a continuity decision in every materially adverse case. It need not guarantee a stay. The reviewer should assess urgency, apparent merit, harm to the holder, harm to others, security risk, reversibility and public interest. The institution may proceed where delay creates greater danger. What should disappear is silence, because silence usually operates as a decision in favour of immediate effect.

A pause must preserve, not prejudge

Registries have legitimate reasons to resist automatic stays. A fraudulent actor could use review to retain control. A transfer dispute could prejudice an innocent counterparty. Security-sensitive services may require rapid action. Unpaid fees and abandoned records cannot remain unresolved forever. A blanket freeze would convert due process into a tool of obstruction.

The answer is to define preservation narrowly. The registry may maintain the current registration entry while preventing transfer. It may keep essential services active while restricting non-essential changes. It may require disputed money to be paid into escrow, as the RIPE NCC rule contemplates for an invoice. It may demand current security contacts, prohibit asset movement or require cooperation with an urgent investigation.

The applicant should make a concise showing: a reviewable decision, an arguable ground, a risk of non-compensable harm and willingness to comply with protective conditions. The institution should respond within a short fixed period. Refusal should include reasons and be reviewable by an independent person before the act becomes irreversible where practicable.

This design separates continuity from merits. Preserving a record does not establish ownership or compliance. Maintaining service for a fortnight does not approve the disputed conduct. It merely protects the subject of review. Courts understand this distinction when they issue temporary relief; registries can build the same discipline into their administrative layer and reduce the number of emergency court applications.

Deadlines create hidden constitutional choices

The published routes use different clocks. AFRINIC's appeal clause specifies 21 days for notice. ARIN's request appeal gives 30 business days after the internal escalation outcome. The RIPE NCC procedure uses a reasonable period capped at one year after the dispute begins, following an attempt at direct resolution. APNIC's by-laws set short stages for appointing representatives and conducting good-faith settlement once a dispute notice is given.

None of these periods is inherently correct for every claim. A short limit can preserve evidence and deliver certainty. It can also expire while an operator seeks reasons, authorisation, translation or counsel. A longer period permits preparation but leaves counterparties uncertain. The quality of a deadline depends on when it starts, what notice was given, whether the record is available and whether services remain stable.

A baseline should make the clock conditional on adequate notice. Time should begin only when the affected party receives the decision, reasons, operative rule, effective date and review instructions. Material record disclosure should occur promptly. If the institution corrects or changes its reasons, the period should restart for the changed ground. A good-faith request for an accessible format or authorised representative should not consume the remedy.

Urgent relief needs a separate clock measured in hours or days. Ordinary merits deadlines cannot answer a threat scheduled for tomorrow. Every registry should publish an emergency contact, required showing, decision time and escalation point. Predictability protects the institution as much as the claimant by channelling genuine emergencies away from improvised executive correspondence.

Cost rules select which doctrine gets tested

The five routes also distribute cost differently. APNIC's by-laws generally leave each party with its own costs and split arbitral charges unless the decision says otherwise. ARIN's agreement has each party bear its lawyers while the initiator initially pays arbitral expenses. The RIPE NCC specialist procedure caps defined reasonable procedural costs below EUR 5,000 and places them on the losing party. LACNIC's three-arbitrator route necessarily refers the reader to an external institution's rules and charges for the full picture. Court proceedings add local counsel, filing, evidence and potential cost-shifting.

These differences shape precedent. A large holder can finance a case that clarifies a transfer rule. A small network may surrender even when the issue is equally important. The resulting body of decisions then overrepresents disputes with high asset value or well-funded parties. Apparent legal certainty is partly a selection effect.

Affordable review does not require free proceedings. It requires cost visibility, proportionality and protection against exclusion. The notice should identify expected fees and cost exposure. A specialist first stage should narrow facts before commercial arbitration. Small entities should be able to seek waiver, staged payment or a capped contribution based on published criteria. Bad-faith conduct can still attract costs.

The objective is institutional: recurring policy questions should be tested on merit rather than only when a claimant can fund the forum. A global registry system that depends on consistent confidence should treat review capacity as part of its operating infrastructure, not as a private luxury purchased after an adverse act.

The decisional record is the cheapest form of legal aid

Registry staff usually possess the organised record: application tickets, account history, internal checks, policy references, communications, technical logs and the chronology of escalation. The affected party has part of this material but may not know which fact or interpretation was decisive. Without disclosure, it must guess and plead broadly.

That asymmetry raises cost and weakens review. Lawyers spend time reconstructing events. The claimant challenges possibilities rather than findings. The institution then answers that the case is vague, although it holds the information that would make it precise. An independent reviewer receives two narratives instead of one indexed record.

Every material adverse decision should therefore generate a standard file. It should contain the request or issue, authority, policy version, evidence relied upon, material correspondence, findings, reasons, decision time, effective action and prior escalations. Withheld material should be listed by category and legal or security basis. The reviewer must be able to inspect confidential material even when the claimant cannot receive it in full.

ICP-2's emphasis on record keeping and auditability supports this approach without dictating a litigation rule. A registry that cannot assemble the decisional spine cannot demonstrate neutral administration. Early disclosure lowers the cost of meritorious claims, exposes weak claims sooner and helps the institution correct factual error before positions harden. It is the rare reform that improves access, speed and defence quality simultaneously.

Scope is the first point at which remedies vanish

An operator may see the word "appeal" and reasonably assume that a serious adverse act is covered. The text may say otherwise. ARIN's named appeal is directed to number-resource request decisions. The RIPE NCC procedure lists covered relationships and excludes some services with their own court clauses. AFRINIC's right is worded around the performance of an assigning registry. APNIC's management form concerns operational decisions, while its by-laws define a broader class of institutional disputes. LACNIC's agreement addresses controversies between contracting parties.

These boundaries are legitimate if they are visible and if another route fills the gap. Trouble arises when an act falls between categories: too operational for a policy appeal, too administrative for commercial arbitration, too urgent for membership governance and too contract-specific for a general complaint. The claimant circles the institution while the decision takes effect.

A baseline should use a residual clause. Any materially adverse exercise of registry authority affecting registration or associated essential services must have an identified independent review route, unless a court or statutory tribunal already has exclusive jurisdiction. The institution may assign different specialists to different subjects, but it cannot answer a scope objection without naming the competent alternative.

This is not an invitation to challenge every help-desk exchange. Materiality can be defined by denial, suspension, termination, revocation, reissue, transfer refusal, significant record change, certification restriction or another action with comparable effect. Minor service complaints remain in ordinary channels. The residual clause prevents serious cases from disappearing through taxonomy.

Independence is a chain, not a label

A reviewer can be independent in title but dependent in appointment, information or remedy. Conversely, an internal reviewer can exercise genuine professional distance and correct many errors quickly. The right question is not whether the first stage sits inside the registry. It is whether the overall chain includes a decision-maker free to disagree, obtain the record, give reasons and order an effective result.

Board review has democratic legitimacy but may involve institutional self-interest. Management escalation has expertise but hierarchical loyalty. Community arbiters bring sector knowledge but may share professional networks with the parties. Commercial arbitrators bring formal independence but may lack number-policy experience. Courts bring public authority but depend on admissible claims and national procedure.

Appointment rules should disclose conflicts, provide a method of objection and avoid case selection by the official whose decision is challenged. Tenure and remuneration should not depend on outcomes. Clerical support can come from the registry if the reviewer controls it and confidential submissions are protected. External legal advice should be available where the institution is a party, as the RIPE NCC procedure recognises.

Most importantly, the independent stage must possess a remedy. A beautifully reasoned recommendation that arrives after reissue is not equivalent to an order capable of preserving and correcting the record. Where the internal reviewer lacks coercive authority, the rules should provide fast recognition by the board or a direct path to arbitration or court.

Publication determines whether review improves the institution

Private resolution can protect commercial data and encourage settlement. It can also conceal inconsistent treatment. APNIC's arbitral confidentiality, the publicity provisions in the RIPE NCC specialist procedure, and the relative silence of other concise clauses reveal different choices about what the community learns.

A useful publication rule separates identity from doctrine. The institution should publish the issue, policy version, material facts at an appropriate level, reasoning, remedy, duration and cost outcome. Customer names, network security detail, personal data and confidential commercial evidence can be removed unless disclosure is necessary and lawful. Settlements can produce a short statement of the institutional question without exposing negotiated terms.

The aim is not spectacle. Published reasons reduce repetitive disputes, guide staff and let members test whether similarly situated parties receive similar treatment. They also reveal scope gaps. If every difficult case settles privately, the public rules remain deceptively simple while the institution accumulates unpublished exceptions.

Aggregate reporting should include rejected and abandoned claims, not only completed appeals. The number of requests for a pause, the rate granted, time to record disclosure, cost, outcome and reason for withdrawal can show whether access works. An empty docket may indicate excellent decisions; it may also indicate that the route is obscure or unaffordable. Publication lets members distinguish the two.

Membership accountability cannot depend on litigation alone

The registries are rooted in member and community structures, but a contractual dispute can become detached from collective oversight. Confidential arbitration resolves the parties' rights. It does not necessarily tell members whether a policy is unclear, an appeal is too expensive or a service can be interrupted before review.

Boards should receive independent annual reporting on the remedial system, while avoiding interference in individual merits. Members should be able to amend procedures, approve budgets for independent review and question recurrent delay. APNIC's strong member power over Executive Council decisions illustrates the constitutional capacity of membership, even though the threshold makes it unsuitable as an individual remedy. Similar governance levers can oversee the design rather than retry each case.

The separation protects both sides. Claimants should not have to campaign publicly to obtain relief. Registries should not face a membership referendum on confidential facts. An independent case route resolves the dispute; aggregate democratic oversight repairs the rules.

Metrics should be reported by dispute type and stage: request denial, transfer, suspension, billing, record accuracy, certification or termination. The report should show which body decided, whether continuity was preserved and whether the institution accepted the result. If a board repeatedly overrides or delays reviewer recommendations, members can see that the apparent route lacks force.

Cross-regional transactions expose the baseline gap

Inter-RIR transfers and multinational networks make remedial divergence harder to contain. A transaction can involve a source holder in one region, a recipient in another, brokers, lenders and customers elsewhere. Each registry applies its policies and due diligence, but a dispute may turn on different notices, timelines and review routes at each end.

Suppose one registry pauses a contested transfer while another treats its decision as immediately effective. The parties may face inconsistent records or contractual uncertainty even if each institution follows its own rules. Suppose one end publishes reasons while the other keeps arbitration confidential. The market sees half of the explanation. Suppose one route is affordable and the other requires a distant three-arbitrator proceeding. The less accessible end can determine whether the whole transaction survives.

A minimum baseline would not merge the cases. It would provide interoperable procedural signals: notice that review is pending, a common description of the preserved state, named contacts, aligned deadlines where possible, and a rule against irreversible reissue while a competent stay is active. Registries could exchange only the information required for coordination and protect confidential evidence.

The system already coordinates technical uniqueness. Coordinating the status of a disputed inter-regional transaction is an extension of that operational discipline. Without it, regional procedural autonomy can produce global ambiguity, the very outcome a registry system is meant to reduce.

Scarcity turned procedure into market structure

In the early years after 2000, many disputes centred on eligibility and policy application in a world where additional address space could still be allocated more readily. IPv4 exhaustion changed the economic setting. Existing blocks gained transaction value, transfer policy became more important and registry status became a material input to financing and acquisition.

Procedure now affects bargaining power. A party with a credible route to a rapid hold can negotiate without fearing immediate loss. A party facing expensive arbitration and no stated pause may accept a discount or settlement unrelated to the merits. A registry's remedial constitution therefore influences the transfer market even when the registry takes no position on price.

This does not mean every address dispute is a property contest or that scarcity invalidates policy controls. It means the cost of administrative error has risen. Fraud detection, sanctions compliance and due diligence are necessary precisely because the assets and services around registration attract strategic behaviour. Strong controls require strong review, since a false positive can be expensive and a weak remedy can encourage pressure through complaints or litigation.

The baseline is thus pro-administration. Clear reasons and independent review make decisive enforcement more credible. Temporary preservation can be denied where evidence shows fraud or acute security risk, but the denial is recorded and reviewable. Markets receive better signals because institutional action is both firm and procedurally reliable.

Court access should be designed, not merely acknowledged

Every registry operates within national law. A contract cannot make courts irrelevant, and an internal panel cannot issue every coercive remedy. Yet a generic statement that parties may go to court is not an adequate continuity plan.

The useful questions are operational. Which court can hear an urgent application? Must arbitration start first? What notice is required? Can the court preserve the status quo? Is a foreign member permitted a local alternative? What evidence will the registry provide? Will it agree not to argue that the claim is premature after directing the party to that forum? How will an order reach staff who control the relevant service?

ARIN's agreement answers part of this by naming courts and the purpose of status-quo relief. APNIC's by-laws preserve interlocutory proceedings despite the preliminary dispute sequence. The RIPE NCC procedure links non-compliance with a short opportunity to submit the dispute to a competent court. These are different bridges, but each recognises that urgent judicial relief and ordinary merits resolution can coexist.

All five institutions should publish a court-readiness protocol without offering legal advice. It should identify service addresses, authorised contacts, accepted electronic delivery, the process for authenticating records, and an emergency implementation team. The protocol would not concede jurisdiction or merits. It would prevent a valid order from losing effect because the institution cannot route it internally before a scheduled action.

Ombuds-style review is valuable at the seam

There is a useful role between customer service and adjudication. An independent ombuds-style office could help an operator understand scope, obtain the decision file, identify the correct respondent, mediate a narrow correction and refer urgent cases to the person authorised to pause action. It could also collect patterns that formal tribunals see only case by case.

Such an office should not become a detour. Filing with it should toll relevant internal deadlines, and urgent referral should not require completion of mediation. Its recommendations should be published in aggregate, while confidential facts remain protected. The office should report to the membership or an independent committee rather than the executive responsible for the disputed service.

The five registries need not use the same title. Some may assign the function to an independent complaints officer, a standing reviewer or a panel secretariat. The baseline concerns capability: navigation, record access, maladministration review, mediation, systemic reporting and emergency referral.

It also concerns limits. An ombuds officer should not finally determine disputed contractual rights without authority and safeguards. Nor should an institution cite a helpful conversation as proof that independent review was available. The office adds a low-cost seam to the remedial chain; it does not replace the link to a binding forum.

The minimum baseline can be stated in ten duties

First, every material adverse decision should identify the act, authority, policy or contract term, material facts, reasons, effective time and responsible official. Second, the notice should name every available internal and external review route, deadline, expected cost and emergency contact. Third, the institution should provide the indexed decisional record promptly, with explained and reviewable withholding.

Fourth, a person not involved in the original decision should conduct a rapid reconsideration capable of correcting obvious error. Fifth, a materially affected party should have access to an independent merits reviewer with power to obtain evidence, give reasons and direct an effective remedy. Sixth, the reviewer should decide temporary preservation under published criteria before irreversible action where practicable.

Seventh, cost should be predictable and proportionate, with waivers, caps or staged contributions where ordinary charges would prevent a serious claim. Eighth, time should run from adequate notice and pause when the institution has not supplied the required record. Ninth, anonymised or appropriately redacted outcomes and aggregate access metrics should be published. Tenth, a clear route to urgent court or binding arbitral relief should exist where the internal body lacks coercive authority.

These duties do not select a winner, a national law or one institutional model. ARIN could retain its contractual arbitration, APNIC its corporate route, RIPE NCC its specialist panel, LACNIC its Uruguayan conciliation and AFRINIC its rapid appeal. Each would add or clarify the elements needed to make the route practically equivalent at the baseline.

Common forms would make regional diversity legible

Implementation need not begin with a treaty. The five registries could adopt common fields for adverse notices, review requests and continuity decisions. The substance would remain local, but an operator working across regions would recognise the structure.

The adverse notice would identify decision type, affected services, policy version, evidence categories, effective time and available routes. The review form would ask for the challenged finding, requested remedy, urgency, continuity risk and confidential material. The preservation decision would address arguability, reversibility, harm, security, third parties and conditions. A final decision would separate facts, policy interpretation, remedy and implementation.

Common forms would also improve comparative data. The registries could report median time to reasons, record delivery, internal correction, independent decision and implementation. They could compare the share of cases in which a pause was sought or granted, without exposing member identities. Divergence would become a subject for deliberate explanation rather than accidental inheritance.

Uniform forms are not uniform law. The same field for governing forum might contain Virginia arbitration, Queensland proceedings, Dutch court review, Uruguayan arbitration or Mauritian remedies. That visibility is the point. An interoperable system can preserve regional authority while making critical procedural differences explicit before they cause harm.

Mutual recognition should be limited to procedural status

A more ambitious baseline could require each registry to recognise certain procedural states issued by another: review pending, temporary hold active, final decision issued, appeal period open. Recognition would not mean accepting the other registry's interpretation of allocation policy or national law.

The narrow purpose is to prevent conflicting operational action. If a source registry has stayed a transfer, the recipient registry should not complete the corresponding change until the status is clarified. If a court order binds one registry, the others should have a contact route to assess whether coordinated records require temporary protection. If a final decision changes the authorised party, implementation should be synchronised.

Safeguards are essential. The notice must be authenticated, identify authority and duration, and disclose enough to show scope. A registry may refuse recognition where it would violate local law, security duties or a binding order, but it should give reasons and notify the other institution promptly. Temporary recognition should expire unless renewed.

This is administrative comity, not a supranational court. It reflects the same practical logic used to coordinate resource transfers and uniqueness. The registries remain responsible to their own members and laws; they simply avoid making a cross-regional dispute irreversible while competent review is genuinely under way.

Objections to a baseline deserve serious answers

The first objection is abuse. More review and easier stays could encourage strategic delay. The answer is a standing threshold, short emergency submissions, conditions, security where proportionate, and costs for demonstrated bad faith. Direct screening of conduct is better than denying all small operators a meaningful route.

The second objection is legal diversity. Five institutions sit in different jurisdictions and use different corporate forms. That prevents a single appellate court, not a shared set of administrative duties. Notice, reasons, record access, conflict rules and a continuity assessment can be implemented under many legal systems.

The third objection is cost. Independent review, publication and an ombuds office require funding. So do opaque disputes: executive time, outside counsel, emergency litigation, inconsistent records and reputational damage. A specialist first stage can reduce total cost by narrowing issues before arbitration or court.

The fourth objection is institutional autonomy. Regional communities built their own processes and should not be governed by a remote centre. The baseline should be adopted and audited collectively by the five communities, with local implementation. It protects autonomy by making regional administration more defensible, not by transferring merits decisions elsewhere.

The fifth objection is low case volume. Rare disputes may not justify permanent machinery. A shared roster, part-time ombuds capacity and common forms can scale. Low frequency is also a reason to standardise: neither staff nor claimants should invent emergency procedure for a case that appears once in several years.

What evidence would change this assessment

This comparison relies on published technical standards, recognition criteria, the NRO matrix and the dispute instruments to which it points, with closer attention to the ARIN and RIPE NCC mechanisms. It does not claim that every informal accommodation, court remedy or internal practice appears in those documents.

The assessment would change if the registries published evidence that materially equivalent safeguards already operate across all five regions. Relevant evidence would include complete adverse-notice templates, independent appointment records, fee schedules, waiver outcomes, stay criteria, median timelines, anonymised rulings and proof that urgent orders are implemented before disputed services change.

It would also change if empirical data showed that the differences do not affect access or continuity. If small and large operators reach independent review at similar rates, if records arrive promptly, if emergency applications are rare because internal holds work, and if comparable disputes produce comparable procedural treatment, the case for additional common rules would weaken.

Conversely, abandoned appeals, expired deadlines, unreviewed scope disputes, unaffordable arbitration or corrected decisions after irreversible action would strengthen it. The important evidence is not the number of procedures on a governance page. It is the journey from adverse notice to effective remedy.

The baseline must not become a global merits appeal

There is a temptation to solve divergence by creating a central tribunal above the five registries. That would introduce its own problems: distance from regional policy, uncertain legal authority, appointment politics, cost and the risk of confusing technical coordination with global adjudication.

The minimum baseline proposed here is deliberately less ambitious. It establishes conditions for a fair and effective route, then lets regional institutions and national courts decide substance. A central body might maintain common forms, publish comparative statistics, coordinate cross-regional status and audit whether the duties exist. It should not routinely decide whether a particular member satisfied a local allocation policy.

Nor should the baseline create an appeal against every policy choice. Open policy development, board accountability and case review address different questions. A reviewer can test whether an adopted rule was applied correctly and lawfully without substituting personal policy preferences. Members can change the rule prospectively through regional governance.

This boundary preserves legitimacy. The global system protects interoperability and minimum procedural reliability. Regions retain their authority over policy and institutional design. Courts retain their authority under law. Claimants receive a meaningful route rather than a new distant layer that may duplicate existing forums.

From accidental geography to deliberate equivalence

The five-registry model was built to combine global coordination with regional participation. Its success should not be measured only by whether allocations remain unique and records are served. It should also be measured by what happens when the institution may be wrong.

The present arrangements contain valuable components. ARIN names a status-quo court route. APNIC combines operational escalation, member authority and arbitration. The RIPE NCC provides a specialist panel, cost ceiling and structured transition. LACNIC places conciliation before arbitration. AFRINIC sets a short written appeal timetable and a pre-termination opportunity to respond. None of these features should be discarded merely for uniformity.

But their combination across five separate constitutions leaves uneven access. The same class of globally interoperable resource can be protected by different deadlines, costs, reviewers and pause mechanisms because of the region in which the holder qualified for registration. That is not meaningful local policy diversity. It is accidental remedial inequality.

The answer is deliberate equivalence: common duties of notice, reasons, record, independent review, affordability, temporary preservation, publication and external enforceability. Each registry can express those duties through its own law and community. Operators would still need to prove their cases. Registries would still act against breach, fraud and non-payment. What would change is the first verdict. It would be delivered on evidence, not by registration geography.