A number-resource dispute rarely announces itself as a technical failure. Packets may still move. Customers may still reach services. Reverse DNS may still answer. Route-origin attestations may still validate. The weakness appears elsewhere, in the room where a buyer, lender, customer, lessor, court or board asks a colder question: if the status of this IPv4 block is contested, who can decide what the registry records, how quickly, on what evidence, with what interim protection and at whose expense while everyone waits?
That question is not procedural ornament. It is part of the asset. IPv4 scarcity has made registry recognition economically important even where the legal vocabulary avoids calling addresses property. A block that can support customer contracts, corporate restructuring, financing, leasing, acquisition planning or a clean sale is worth more than an identical block whose registry position can be trapped by a badly designed dispute. The difference lies not in the addresses themselves but in the credibility of the forum standing behind the ledger.
AFRINIC is a useful test because its recent history has made the forum visible. Public reporting and litigation records have described alleged manipulation of African IPv4 records involving dormant or defunct organisations, a large dispute with Cloud Innovation over resource holdings and use conditions, litigation that affected AFRINIC's financial capacity, a long period without ordinary board stability, receivership under Mauritian court supervision, an annulled election after questions about voting authority and powers of attorney, a later attempt to restore board function and continuing conflict around the institution's future. Those episodes do not settle the merits of any individual claim. They do show why dispute resolution is infrastructure. When the body maintaining the ledger also becomes a participant in the dispute, the design of the forum becomes part of the value being fought over.
This is not the same subject as abuse-contact enforcement, identity validation or appeals after an adverse decision. Those questions matter, but they sit inside a wider economic problem. Dispute-resolution economics asks how a scarce-resource ledger should convert contested claims into credible, bounded, reviewable records. It asks how a registry can protect the ledger without becoming a discretionary market regulator, how a resource holder can challenge registry action without threatening institutional paralysis, and how non-parties can rely on the record while the fight continues.
The market audience is larger than the two names on the file. A contested transfer can affect a buyer's closing conditions, a seller's liquidity, a lender's covenant package, a customer's migration plan, a leasing counterparty's revenue, an upstream provider's risk assessment, an auditor's valuation and a court's willingness to preserve the status quo. Smaller members also watch. They want a registry strong enough to resist fraud and weak enough not to make compliance conflicts existential. A slow, opaque or visibly self-interested process does not merely inconvenience litigants. It adds a risk premium to every resource administered under the same institutional design.
The central question, then, is not whether AFRINIC should prevail against any particular critic or holder. It is whether the dispute system can survive the possibility that the registry loses, the holder loses, or both lose on different points, while the market still trusts the ledger. Mature market institutions are not those without serious disputes. They are those whose disputes are contained, priced and resolved without making every file a referendum on the institution's survival.
The price of uncertainty enters before the dispute begins
Two IPv4 blocks can be technically indistinguishable and economically different. They may be the same size, routed with equal reliability, attached to similar customers and supported by similar operating teams. Yet one can command a better price because its registry status is legible and its conflict path is predictable. The other can attract a discount because counterparties do not know what happens if an old claimant appears, a transfer is challenged, a court order arrives, the registry raises a policy objection or the holder and registry disagree over permissible use.
Markets price that uncertainty long before anyone files papers. A buyer asks for indemnities. A broker extends diligence. A bank excludes address-dependent cash flow from a borrowing base. A lessor raises margin. A customer demands migration rights. A small seller accepts a lower price because it cannot wait for months of clarification. The discount is not simply the expected loss of the case. It is the cost of not knowing which forum matters, what facts matter, what interim protections exist and whether ordinary operations continue while the matter is tested.
Scarcity makes this discount visible. IPv4 resources now sit inside business plans, customer contracts, hosting revenue, merger models and leasing arrangements. A dispute over registry recognition is therefore a dispute over expected cash flow. The public address record has become more than a directory entry. It is a reliance instrument used by parties who may never attend a registry meeting and may have no appetite for internet-governance theory.
AFRINIC's controversies illustrate the point. Allegations of historical address-record manipulation suggested that stale entities, dormant contacts and registry knowledge could be monetised if controls failed. The Cloud Innovation dispute showed how a disagreement over large holdings could spill into court and affect the registry's finances. Receivership and contested elections showed that the authority of the registry itself could become uncertain. Each episode widened the market's question from "who is recorded against this prefix?" to "what happens if the institution answering that question is also under stress?"
A dispute marker can help or harm depending on how precise it is. A narrow marker saying that a proposed transfer is paused because a named claimant has produced specified evidence may preserve value. It tells outsiders that the last verified state remains in place while one transaction is examined. A broad marker that leaves outsiders guessing whether the problem is fraud, unpaid fees, litigation posture, policy interpretation or institutional hostility does the opposite. It stains the file without pricing the risk.
Silence can be just as costly. If a dispute is pending and the registry will not say what is frozen, what remains available, what evidence category is being assessed or what path will decide the matter, conservative counterparties assume the worst. Credit committees and acquisition teams are not rewarded for optimism. They are rewarded for reducing uncertainty. A registry lowers the cost of capital by classifying disputes and explaining operational consequences. It raises the cost by treating confidentiality as a reason to leave every outsider in fog.
The first economic function of a dispute system is classification. It should distinguish a rival claimant from a transfer-document defect, a court order from a policy disagreement, suspected forgery from stale contact data, a payment dispute from a resource-use dispute, a registry enforcement action from a private fight between members, and a corporate-control dispute from an address-ledger question. Without classification, every contested file becomes an undifferentiated risk. In a market, undifferentiated risk is expensive.
A registry ledger is a market institution
Regional internet registries are often described as technical coordination bodies. That description captures much of their daily work and little of their economic significance during conflict. A registry does not need to own the addresses to affect asset-like value. It does not need to run an exchange to affect liquidity. It does not need to operate a bank to affect credit. It only needs to decide whether the public ledger recognises a holder, a successor, a transfer, a dispute flag, a suspension, a court restraint or a remedy.
The value of the ledger lies in its ordinariness. In good times, a registry record is dull. It lets others rely on uniqueness, contact data, delegation history and operational continuity without negotiating the politics behind each entry. Disputes make dullness hard. When a holder and a rival claimant disagree, the bookkeeper is asked what the book should say. When the holder and registry disagree, the bookkeeper is also a party with budget exposure, legal strategy, reputational interest and institutional incentives.
That role conflict cannot be wished away. A registry must be able to prevent forged transfers, reject defective requests, comply with lawful orders, protect evidence, maintain uniqueness and stop corruption of the ledger. A powerless registry would be unsafe. But a registry that can impose high-consequence outcomes without separation between ledger maintenance and institutional self-interest is unsafe in another direction. The danger is not registry authority itself. It is unpriced, unchecked and self-interested authority over a scarce-resource record.
AFRINIC's position sharpens the problem because institutional continuity and resource disputes have overlapped. The registry has had to defend its own functioning while making or defending decisions that affect valuable holdings. Resource-holder interests and advocacy groups have attacked aspects of registry power. Coordination bodies and peer institutions have expressed concern about continuity of the RIR system. Each side points to real risks. A registry cannot be disabled by one determined litigant. A holder cannot be left without meaningful recourse when registry action threatens a valuable operating position. Forum design must account for both incentives instead of pretending one of them is imaginary.
The bookkeeper-judge conflict is most acute when the requested remedy would alter recognition itself. If AFRINIC believes a record is false, a transfer is forged or a holder has breached a defined condition, it needs tools to protect the ledger. But if the proposed consequence is termination, reclaim, transfer blockage, refusal to recognise a successor, or a public status that destroys liquidity, the registry's own view should not be the only check. A credible system asks whether the last verified record can be preserved while the claim is tested, whether a narrower remedy would protect the ledger, and whether an independent decision-maker should review the evidence before irreversible harm occurs.
Communications also have economic weight. A registry statement about a dispute can move counterparties even if the statement is not a final decision. If a notice says or implies that a certain business model, transfer request or holder status is unacceptable, buyers and lenders react. The statement may be accurate, contested, incomplete or advocacy. A neutral registry separates factual status notices from litigation argument. It avoids making its preferred interpretation look like the final condition of the record before the forum has acted.
Bookkeeping and judging can coexist only through role separation. Staff can maintain records. A dispute office can classify the issue and preserve the last verified state. An independent reviewer or panel can decide high-consequence contested matters. Courts can handle corporate control, fraud, insolvency and damages. The board can set policy and budgets but should not be the first and last judge in disputes where the institution's own authority, liability or litigation posture is at stake. Separation is not bureaucracy for its own sake. It is what lets the registry act without making every action look like self-dealing.
Dispute resolution is a market utility, not courtroom theatre
Public conflicts invite dramatic storytelling. One side claims to defend a regional public interest. Another says it is resisting a monopoly gatekeeper. One side warns that a large holder can exhaust a registry. Another warns that an unchecked registry can destroy reliance built over years. Courts issue interim orders. Campaign sites publish documents. Conferences and statements supply language for supporters. The drama may be unavoidable, but it is not the central economic issue.
The central issue is that dispute resolution supplies a market utility. It converts competing claims into a credible state of the ledger. It tells non-parties what they may rely on while the merits are examined. It allows a buyer to close or defer with defined risk, a lender to decide whether collateral-like reliance is usable, a customer to know whether migration is prudent, and a court to preserve the subject matter without guessing how registry operations work. The value lies in containing uncertainty before it contaminates adjacent relationships.
A market utility needs predictable entry points. A holder should know where to challenge a registry action. A rival claimant should know how to submit evidence without freezing a record by accusation alone. A buyer should know whether a pending claim blocks only the transfer or affects routine maintenance. A court should know how an order will be received, verified and reflected in registry status. A member should know whether a dispute over voting authority has any connection to resource authority. Without defined entry points, disputes become contests over access to staff, lawyers, board members, courts and public opinion.
It also needs predictable outputs. A well-designed forum should be able to say that the record remains unchanged, that only a transfer is paused, that a conflict marker has been added, that routine services continue, that publication and delegation services continue, that a challenged representative cannot issue new instructions, that a court order restrains a specific action, that an independent review will decide a preservation request by a defined date, or that a question is outside registry competence and belongs in court. These outputs are not exciting. They are the units of market confidence.
AFRINIC's receivership shows the difference between institutional rescue and ordinary dispute utility. A receiver can preserve business value, maintain the status quo and help restore corporate organs under court supervision. That kind of intervention may be necessary when governance fails. It is not, however, a permanent market forum for transfer disputes, authority conflicts, policy disagreements, leasing controversies or high-consequence registry actions. Emergency plumbing keeps the building standing. It does not by itself design the traffic rules inside.
If the only credible forum is a court, the system is too blunt. Courts are essential when legal rights, fraud, insolvency, damages, corporate authority or public orders are genuinely at issue. They can compel evidence, bind parties and supervise remedies. But courts are expensive, slow and designed for broader legal controversies. They are not built to decide, as a routine matter, how a conflict marker should be worded, whether reverse-DNS delegation should continue during a transfer pause, whether an RPKI publication service should remain stable, or what proof is enough to narrow a stale-record claim. Those registry-specific questions should be answered before courts are forced to carry the whole burden.
The point is not to keep judges out. It is to ensure that by the time a court is needed, the registry layer has preserved continuity, classified the dispute and made the technical consequences legible. When those tasks are absent, parties seek extreme remedies because mild ones do not exist. A bank-account freeze, emergency injunction, winding-up application or external intervention can then become the first effective tool rather than the last. That is not proof that courts are the enemy of registry governance. It is evidence that the ordinary forum is missing.
Forum design decides who can afford to be right
A dispute forum is not neutral merely because it is theoretically open to all members. The cost of using it determines who can afford to be right. A large holder with counsel in several jurisdictions can file motions, prepare expert reports, absorb delay and use publicity. A small ISP cannot. A public university, government network or regional provider may be constrained by procurement rules, budget cycles and internal approvals. A buyer of a modest block may abandon a transaction rather than spend more on the forum than the transaction is worth.
This matters in AFRINIC's service region because many operators depend heavily on relatively small allocations. A network may hold one ASN and a modest IPv4 block that is vital to customers but too small to justify prolonged litigation. If the dispute path is expensive, unclear or perceived as biased, the operator's practical options shrink. It may accept an adverse status, avoid correcting records, abandon a transfer, accept a lower price or sign unfavourable terms with a better-resourced counterparty. A formal right to contest has little economic value if contesting costs more than losing.
Forum design therefore redistributes bargaining power. A process that requires Mauritian litigation for routine registry disagreements favours parties that can litigate there. A process that demands extensive documentary production before narrowing the issue favours actors with strong archives and lawyers. A process with no interim preservation favours the party that can survive uncertainty. A process that lets mere allegations pause transfers favours strategic objectors. A process that lets registry enforcement continue during review favours the registry. These are not technical details. They allocate value.
The same principle applies to expertise, language, evidence format and timing. A forum that expects all parties to understand RIR policy history, Mauritian company law, IPv4 transfer practice, routing-security implications, corporate authority and litigation posture creates a high threshold. Some complexity is unavoidable. The answer is not to reduce every dispute to a checklist. It is staged design: classify first, identify the narrow fact in issue, require evidence relevant to that fact, preserve what can be preserved, and escalate only when the matter truly demands a heavier process.
For AFRINIC, the legitimacy problem is as important as the transaction cost. An expensive or captured forum will not settle disputes. It will export them into louder arenas. The losing side will argue that the process was designed to exhaust it. The registry will describe repeated challenges as obstruction. Smaller members will see instability without being sure which claim is correct. A narrower, cheaper, staged process reduces the political temperature because fewer disputes need to become existential tests.
Good forums do not guarantee satisfaction. They make defeat credible. A holder can accept losing a transfer challenge if the reason is specific, the standard known, the reviewer separate from the original decision and the consequence proportionate. A registry can accept losing an enforcement dispute if the decision preserves the ledger and clarifies the boundary of its authority rather than humiliating the institution. A buyer can accept delay if the deadline and closing condition are clear. Credible defeat lowers the expected cost of future conflict.
This is why dispute rules should be designed before the next crisis, not during it. Once a major dispute begins, every reform looks factional. A stay rule becomes a gift to one side. A stricter evidentiary threshold becomes a weapon for another. An independent panel becomes a way to bypass an unfriendly board. Neutrality needs rules known before the next unpopular claimant, powerful registry officer or strategic objector appears. Ex ante design is the cheapest form of institutional insurance.
Evidence is the forum's currency
Dispute resolution is only as credible as its evidence standard. A registry can call itself neutral, but neutrality has little meaning if some proof is accepted casually, other proof demanded unpredictably or burdens shifted according to the political attractiveness of the party. In a scarce-address market, evidence is currency. It buys recognition, delay, preservation, transfer approval, conflict status or rejection.
AFRINIC's history explains why loose standards are dangerous. Allegations of historical address-record manipulation showed that dormant entities, stale contacts, old correspondence and insider knowledge can have value if the record is not protected. A serious forum must be able to require proof of legal existence, current authority, corporate succession, lack of forgery, court status and connection to the resource. It should not treat portal access, old emails or private contracts as automatically sufficient when the value at stake is high.
The opposite danger is equally real. A registry can demand evidence so broadly that the request becomes a tool of control. A dispute over a transfer may require proof of authority and eligibility. It should not automatically become an audit of every customer, every historical need statement, every geographic deployment and every change in the holder's business model. At that point the forum is no longer asking whether the ledger can safely record a transaction. It is asking whether the holder deserves approval for its commercial life. That is a different institution.
The useful standard is relevance plus proportionality. Relevance asks what fact must be proved for the registry to protect the ledger. Proportionality asks how much proof is justified by the consequence. A rival claimant to an old block may need to show corporate continuity, original connection, current authority and absence of conflicting orders before the registry even pauses the record. A routine buyer may need proof of authority and eligibility but not a general business audit. A registry alleging fraud should carry a stronger file than a registry asking for a missing form. A party seeking revocation should face a higher burden than a party seeking a temporary marker.
Evidence standards also need action categories. "Insufficient evidence" is too vague to support a market decision. Insufficient for what: completing a transfer, changing the holder, lifting a marker, maintaining the last verified state, publishing an allegation, referring the matter to court or revoking recognition? A document may be insufficient to prove a transfer but sufficient to justify a short pause. A court filing may show that a dispute exists without deciding the merits. A suspicious signature may justify confirmation without justifying a public accusation of fraud.
The forum must also handle asymmetry. The registry often controls correspondence files, historical tickets, account records, validation logs and staff interpretations. The holder controls corporate documents, contracts, customer facts and operating history. A rival claimant may control old archives. Courts may hold orders whose registry implications are not obvious to technical staff. A fair process defines what each party can see, what may be redacted, how confidential material is handled and how a decision-maker tests evidence that cannot be published. Confidentiality should not allow one side to use secret proof without a meaningful answer.
The 2025 election controversy offers a parallel, even though it concerned governance rather than resource status. Questions about powers of attorney and voting authority were not merely vote-counting questions. They required standards for who could grant authority, whether the grant was genuine, whether it applied to the relevant act, whether members could verify or revoke it, and what evidence justified annulment. Resource disputes need the same discipline. A power of attorney for a transfer, a board resolution for a merger, a court order affecting a holder and a letter claiming an old allocation must all be tested by standards known before the result is politically convenient.
The strongest evidentiary system is not the harshest. It is the one that lowers the cost of honest proof and raises the cost of strategic ambiguity. Honest holders should know what will satisfy the forum. Fraudsters should not be able to exploit loose records. The registry should not be able to expand the case after every answer. Buyers and lenders should be able to read a status category and understand the exposure. Evidence in a market institution is not paperwork for its own sake. It is the bridge from contested facts to reliable action.
Remedies should fence the dispute rather than contaminate the market
A dispute system is usually judged by who wins. It should also be judged by how much collateral damage its remedies create. In a registry context, the wrong remedy can spread a conflict from one file to customers, unrelated resources, reverse DNS, RPKI publication, public records, bank relationships and member confidence. A good remedy isolates the contested point while preserving everything already verified.
The remedy spectrum is broad. The registry may reject an unsupported complaint, ask for more evidence, mark a transaction as pending, freeze only a proposed transfer, preserve the last recognised holder while blocking a challenged representative, publish a limited conflict status, maintain RDAP, Whois, reverse DNS and RPKI services while a claim is tested, require document escrow, refer questions of corporate ownership or fraud to court, or in severe cases refuse recognition, reclaim resources or end a service relationship. These measures are not interchangeable.
A forged transfer instruction may justify freezing that instruction, not disabling unrelated services. A rival claimant may justify a conflict marker, not immediate substitution of the claimant into the record. A dispute over leasing policy may justify legal or policy review, not sudden disruption of downstream customers. A court order may restrain a specific action, not every service associated with a block. A fee or membership dispute may have consequences, but it should not casually impair security publication if third parties would suffer.
AFRINIC's crisis shows why interim remedies matter. The reported freeze affecting bank accounts during the Cloud Innovation litigation demonstrated how provisional measures can create system-wide risk before merits are settled. Receivership showed how preservation tools can keep an institution alive while also entangling governance repair with resource-market confidence. Election annulment showed that a remedy meant to protect fairness can prolong uncertainty if the reasons and next steps are not trusted. The subjects differ, but the shared lesson is that temporary measures can be economically as important as final outcomes.
For number resources, the starting point should be preservation of the last verified operational state. If a holder was recognised before the dispute, recognition should continue unless strong contrary evidence makes that unsafe and a proportionate remedy is available. If a transfer is disputed, pause the transfer rather than the entire account. If authority is disputed, block the challenged representative rather than all routine maintenance. If a court order is unclear, apply the narrowest safe restraint while seeking clarification. If the registry believes a holder has breached a condition, preserve live services while the breach and remedy are reviewed unless a defined emergency makes preservation unsafe.
This preservation principle protects more than the holder. It protects customers who had no voice in the dispute. It protects other networks that rely on stable records. It protects courts by keeping the subject matter intact. It protects buyers and lenders from avoidable panic. It protects the registry from appearing to use operational continuity as leverage. In a networked market, restraint is not weakness. It is how the forum avoids becoming the source of the harm.
Remedies also need reversibility. A marker can be removed. A transfer pause can expire. A rejected document can be replaced. A temporary hold can be reviewed. By contrast, a public accusation, revoked recognition, disrupted publication service or customer-impacting action is harder to repair. The more irreversible the measure, the stronger the evidence and review should be. That is not procedural elegance. It is expected-value arithmetic. A wrong irreversible remedy can destroy more value than a wrong temporary preservation order.
Symbolic remedies are especially dangerous. A public dispute tempts institutions to demonstrate strength. But a remedy designed to send a message is usually economically crude. If the issue is a bad document, the answer is document control. If the issue is a disputed business model, the answer is a defined policy or legal forum. If the issue is fraud, the answer is to protect the record and involve proper legal authority. Using live resources as a stage for institutional signalling weakens the very ledger the registry is meant to protect.
Delay is a transfer of value
Delay is often described as administrative inconvenience. In a scarce-resource market it is a transfer of value. The party that can wait gains leverage over the party that cannot. The party with cash, lawyers and diversified operations can survive a paused transfer or unresolved status. The smaller party with a financing deadline, customer migration, equipment purchase, acquisition closing or payroll pressure may not. The forum can appear neutral while time quietly redistributes bargaining power.
A buyer waiting for recognition may hold funds in escrow, extend diligence or demand a lower price. A seller waiting for a dispute to clear may lose market timing. A lessor waiting for status clarity may lose customers. A bank waiting for confirmation may withhold credit. A customer waiting for address stability may choose another provider. A court waiting for technical explanation may preserve too much or too little. Each delay creates an option; someone pays for it and someone benefits from it.
AFRINIC's institutional disputes show how delay scales. Years without ordinary board function do not merely affect meetings. They affect budgets, staff morale, resource-processing confidence, legal strategy, election planning and member trust. Receivership can preserve the organisation, but preservation is not the same as predictable decision speed. A restored board may signal capacity, but market participants still ask whether resource disputes, transfer files and membership questions will move through defined timelines or remain vulnerable to broader conflict.
Delay also encourages strategic escalation. If a holder believes the registry will take too long and the asset is at risk, it may seek urgent court relief. If the registry believes a holder will use delay to continue a contested practice, it may seek a stronger interim measure. If a rival claimant knows an objection can block a transfer without an early evidence threshold, the objection becomes a bargaining chip. If a buyer knows a seller needs cash, uncertainty becomes a price cut. Poor forum design teaches all sides to weaponise time.
The solution is not artificial speed. Some disputes are complex. Old records may be messy. Cross-border documents may need translation. Court orders may need clarification. Alleged fraud may require careful preservation. But complexity does not justify open-ended timing. A credible forum should publish expected periods for emergency preservation, initial classification, evidence exchange, transfer disputes, authority disputes, registry-action review, independent review and court referral. The market can price a known schedule. It heavily discounts an indefinite one.
Timing discipline must be paired with interim continuity. A short deadline without preservation can force rushed injustice. Preservation without deadlines creates limbo. The combination matters: preserve the last verified state, classify quickly, require relevant evidence, decide narrow interim effects and escalate with stated milestones. That design reduces the incentive to seek extreme remedies because the ordinary path offers both safety and movement.
Delay costs also shape regional development. Operators that cannot predict registry timing will be cautious about transfers, leasing, mergers and financing. Some will leave unused resources idle because moving them looks risky. Some will avoid formal updates because opening a file might expose them to delay. Some will pay intermediaries because specialists know how to navigate uncertainty. The result is a thinner, more expensive market for address capacity. A system meant to protect the ledger can reduce ledger accuracy if correction and transfer become too slow.
AFRINIC's recovery should therefore be measured partly in days. How long does initial classification take? How soon does a holder receive reasons? How long may a conflict marker remain without an evidentiary upgrade? How quickly can a preservation request reach an independent reviewer? How often do third-party objections pause transfers, and for how long? How often does litigation stop ordinary processing? These metrics are not administrative trivia. They are the time-price of the registry.
Neutrality means the referee cannot control the match and the prize
Neutrality is often reduced to even-handed language. That is too thin. In a dispute over valuable registry recognition, neutrality is structural. The referee should not control the match, define the rules during play, hold the prize, decide the remedy, speak publicly as an advocate and then ask the market to treat the result as disinterested.
This does not mean the registry must withdraw from disputes. It is the only institution able to maintain the record, preserve uniqueness, receive technical updates and apply registry-specific remedies. But neutrality requires the registry to separate functions that create conflicts. The unit maintaining the ledger should not be the same as the body deciding a high-consequence challenge to the registry's own action. The office issuing a factual status notice should not mix it with advocacy about the institution's virtue. The board that has litigation exposure should not be the final evidence tribunal for that litigation.
The practical design is not mysterious. First, a classification function decides what kind of dispute exists and what immediate preservation is needed. Second, a record-maintenance function implements the narrow operational status. Third, an independent reviewer or panel hears contested high-consequence decisions. Fourth, courts handle matters outside registry competence. Fifth, public communications separate status from argument. Each layer has a limited job. The limits are what make the whole system believable.
AFRINIC needs that discipline because the incentives around it are strong. The registry has an institutional interest in survival, authority, budget stability and respect for its decisions. Large holders have commercial interests in transferability, leasing models, litigation leverage and preservation of value. Advocacy groups have reputational interests. Coordination bodies have system-continuity interests. Courts have legal mandates that may not map neatly onto registry operations. None of these interests is illegitimate. None should be confused with neutral resolution.
Funding neutrality matters too. A dispute system financed entirely by general member fees can create moral hazard. The registry may fight because costs are spread across members and downside liability is limited. A large holder may litigate because it can impose costs on the registry and indirectly on everyone else. The answer is not to starve the registry of legal resources. It is to disclose cost categories, choose proportionate remedies and make the ordinary forum cheaper than institutional war. Members should not subsidise avoidable escalation simply because no credible middle forum exists.
Settlement also needs neutral architecture. A registry may fear that settlement signals weakness or validates a contested business model. A holder may fear that settlement gives the registry language to use later against others. Other members may fear that a private bargain changes policy without consent. Those fears are rational when dispute resolution is opaque. Settlement principles should therefore be visible: preserve the ledger, do not secretly rewrite policy, protect third-party continuity, disclose aggregate effects where appropriate and separate private claims from public registry functions. A settlement that restores certainty can be pro-market. A settlement that hides institutional concessions becomes another source of distrust.
The best dispute system makes control of the forum less valuable. If rules are clear, evidence standards known, remedies proportionate and review genuinely independent, neither the registry nor a large holder gains much from procedural domination. They must win on the narrow issue. That is the central economic benefit of neutral design: it lowers the reward for brinkmanship.
Courts are indispensable backstops, not a registry operating model
AFRINIC is incorporated in Mauritius, and Mauritian courts have played a major role in its recent history. That is not surprising. A private membership body with corporate existence, assets, directors, contracts and disputes ultimately sits inside a legal order. Courts can appoint receivers, interpret company law, grant or refuse injunctions, supervise winding-up applications, restrain misleading statements and decide legal claims that a registry forum cannot decide. Rule of law is not an intrusion into registry governance. It is the ground on which private governance stands.
The problem begins when courts become the first effective mechanism for registry questions. If a holder must go to court to preserve resources from an adverse registry action, the registry forum has failed. If the registry requires court receivership to conduct elections and preserve business value, ordinary governance has failed. If outside coordination bodies feel compelled to explain in litigation that number resources should not be treated as ordinary corporate assets available to creditors, the institutional boundary has become dangerously unclear. Courts can rescue. Rescue is costly.
A court-only model is poorly matched to registry granularity. A judge can preserve assets, but may not want to design conflict-status taxonomy. A judge can order or restrain corporate action, but may not decide how RDAP status should be worded during a disputed transfer. A judge can hear allegations of fraud, but may not set everyday evidence standards for legacy-resource regularisation. A judge can appoint a receiver, but cannot permanently substitute for member confidence. The registry needs a technical-economic forum that deals with these narrower matters before they become legal emergencies.
That forum should know its limits. It should not decide who owns a company when that is disputed under corporate law. It should not adjudicate damages for alleged business loss. It should not determine criminal fraud. It should not decide whether a winding-up petition should succeed. It should not reinvent itself as a private court of unlimited jurisdiction. Its job is narrower: decide what the registry should record, preserve or refrain from changing while competent legal institutions decide questions beyond registry function.
This boundary is crucial because public arguments around AFRINIC often bundle different disputes together. A dispute over whether resources are assets of AFRINIC in insolvency is not the same as a dispute over a holder's contractual compliance. A court order correcting corporate authority is not approval of a leasing model. A restraint on public statements is not a final decision about the economics of IPv4 leasing. A receiver's authority to organise elections is not a permanent substitute for member governance. A good registry forum states these distinctions rather than letting every procedural event be marketed as victory on the merits.
Court interaction also needs a protocol. The registry should be able to receive, verify, interpret and implement orders without over-reading them. If an order restrains a transfer, restrain that transfer. If an order preserves the status quo, identify the last verified operational state. If an order is ambiguous, seek clarification rather than selecting the interpretation most helpful to the registry's litigation posture. If a case concerns one holder, unrelated members should not suffer unless the order truly requires system-wide action.
Coordination bodies face a similar discipline. ICANN, the NRO and peer registries have legitimate concern when an RIR's continuity is at risk. Number-resource uniqueness and registry-service continuity are global interests. But continuity concern should not be treated as an endorsement of every registry action or as a decision on every holder dispute. Saying that AFRINIC's function must continue does not prove that a particular remedy is proportionate. Saying that resources should not be available as ordinary corporate assets does not answer every question about holder reliance, transferability or leasing. Emergency continuity language should protect the ledger, not decide adjacent economic controversies.
The hierarchy should be clear: registry-specific classification and preservation first; independent review for high-consequence registry actions; courts for legal rights and remedies beyond registry competence; coordination-body concern only when the registry function itself is at risk. A court-only model reverses that hierarchy and makes each serious resource dispute a threat to the institution.
AFRINIC's history is a stress test, not a morality play
Commentary on AFRINIC often tries to sort actors into heroes and villains. The more useful approach is institutional. What happens when a valuable ledger has allegedly been manipulated? What happens when the registry responds with broad enforcement claims? What happens when a large holder litigates hard enough to affect the registry itself? What happens when the board cannot operate normally? What happens when an election meant to restore legitimacy is itself disputed? What happens when global bodies worry that local corporate proceedings may endanger a public coordination function?
Each question points to dispute-resolution infrastructure. The address-record allegations point to the need for standards strong enough to detect fraud and weak records. The Cloud Innovation dispute points to the need for a forum that separates ledger protection from business-model supervision. The bank-account litigation points to the cost of extreme interim measures. Receivership points to the need for preservation before governance collapses. The election controversy points to authority verification and transparent remedies. Later restoration efforts and continuing litigation point to the difference between formal institutional life and market confidence.
None of these lessons requires adopting AFRINIC's account as the final explanation. Nor does it require treating its critics as neutral observers. AFRINIC has institutional incentives. Cloud Innovation, NRS, LARUS and related commercial interests have legal and economic incentives. ICANN and the NRO have system-continuity incentives. Journalists, analysts and public campaigners work from evidence that is incomplete, contested or strategically presented. The economic method is to ask what each incentive implies and what design would still function if today's sympathetic party became tomorrow's opponent.
That test leads to a narrow conclusion. AFRINIC should be powerful enough to maintain a clean ledger and constrained enough not to act as a discretionary market regulator. It should be able to reject forged records, pause contested transfers, comply with court orders, maintain security publication and preserve services. It should not decide high-value holder disputes through opaque delay, broad status clouds, public advocacy presented as neutral notice or remedies that punish customers before the merits are tested.
The same test applies to large holders. They should be able to challenge registry overreach, obtain preservation, present evidence and demand neutral review. They should not be able to paralyse the registry, use litigation to create leverage unrelated to the merits, hide behind poor records or convert continuity threats into private advantage. A good forum makes both kinds of abuse harder. It gives the registry confidence to act narrowly and gives holders confidence that narrow action will not become arbitrary control.
AFRINIC's future credibility will depend less on winning old arguments than on making future arguments less destructive. A board, budget and strategy can signal recovery. But the market will look for more operational signs: clear dispute categories, preservation rules, evidentiary thresholds, independent review, court-order protocols, conflict markers, timing metrics, cost classification and communication discipline. These are not glamorous reforms. They are the infrastructure that lets a scarce-resource registry become boring again.
The cultural shift may be as important as the procedure. A monopoly-recognition body over valuable operational resources should expect challenges. A challenge does not make the challenger right. It means the institution needs a way to lose some points without losing the ledger. That is what mature market institutions do. They distinguish institutional dignity from institutional immunity.
The risk premium is paid by people outside the dispute
The economic cost of weak dispute resolution does not stop with the litigants. It is distributed across the market as a risk premium. Buyers demand discounts. Sellers lose liquidity. Lenders reduce credit. Customers seek migration rights. Brokers require heavier indemnities. Lessors charge for uncertainty. Smaller operators avoid formal transactions. Investors discount networks whose address dependence sits inside an unpredictable registry environment.
This premium is hard to observe because it appears as deals not done, prices lowered, documents lengthened and time wasted. A clean transaction that never happens is not visible in a public docket. A bank that quietly refuses to value address-dependent revenue does not publish a judgment. A buyer that walks away from an African block because the dispute path looks unclear does not necessarily explain the decision. Yet those private decisions shape market depth.
The risk premium also affects record accuracy. If correcting a record, updating authority or formalising a transfer looks like an invitation to delay, some holders will avoid the process. If stale records seem safer than opening a file, the registry's data quality deteriorates. If specialists become necessary to navigate ordinary interactions, smaller members pay more to participate. The ledger then suffers from the very problem dispute rules were meant to prevent: low trust in formal recognition.
There is a regional development cost. African networks need address resources, capital, customers, interconnection and credible institutions. A dispute system that makes every contested file expensive reduces the ability of unused or underused resources to move toward productive use. It also makes external capital more cautious. Investors can tolerate legal risk if the forum is known and the timeline bounded. They struggle with institutional risk that has no measurable path.
The premium is not eliminated by insisting that number resources are not ordinary property. That statement may be legally important in some contexts, especially where creditors or corporate claimants try to seize assets in ways that threaten uniqueness. But economic value does not wait for property vocabulary. If recognition affects cash flow, transferability, customer retention and financing, then the quality of dispute resolution affects price. A registry can reject property language and still have market power over valuable reliance.
Nor is the premium eliminated by appeals rhetoric. Appeals matter, but a late-stage appeal is not the same as a dispute market. By the time a final adverse decision is appealed, the holder may already have lost customers, a buyer may have withdrawn, a bank may have declined credit and the file may have acquired a public cloud. The market needs earlier tools: classification, preservation, evidence thresholds, narrow interim remedies and timing discipline. Due process after the fact is not a substitute for risk allocation during the dispute.
The premium can be reduced, however, by making status readable. A buyer does not need perfect certainty. It needs to know whether the dispute concerns authority, fraud, payment, policy, transfer eligibility, court restraint or registry enforcement. A lender does not need a promise that no claim will arise. It needs to know whether publication services continue and whether a severe remedy requires independent review. A customer does not need to read litigation filings. It needs to know whether service continuity is exposed. The more readable the status, the lower the discount.
A neutral registry is one that can lose gracefully
Registry neutrality is often described as fairness between parties. In this setting it means something more demanding: the institution can lose a high-consequence dispute without seeming to collapse, and a holder can lose without the result looking like political execution. Graceful loss is the mark of real infrastructure.
For AFRINIC, graceful loss would have several practical features. If the registry loses a review, the decision should restore or preserve the correct record without implying that the registry's existence has been attacked. If a holder loses, the remedy should be tied to the evidence and limited to the harm. If a court narrows the registry's authority, the board should adjust rules rather than describe every limit as sabotage. If a court confirms registry authority, the registry should still choose proportionate remedies rather than claiming a mandate for wider discretion. If a critic is wrong, the registry should answer with records and process, not broad denunciation. If the registry is wrong, correction should not require threats to the whole institution.
This is the economics of confidence. Buyers do not need a world without disputes. They need a world in which disputes are contained. Lenders do not need a promise that no registry action will ever occur. They need evidence that severe action follows known standards and preserves value while reviewed. Operators do not need the registry to approve every business model. They need it to maintain accurate records and define the boundary of its concerns. Courts do not need the registry to be infallible. They need it to keep the subject matter stable and explain its reasoning.
The bargain should be simple. AFRINIC maintains the ledger, protects uniqueness, prevents fraud, records transfers, supports registration data, preserves security and naming services and complies with lawful orders. Members and market participants accept that the registry has a necessary role in contested records. In return, the registry accepts limits: relevant evidence, proportionate remedies, preservation, independent review, timing discipline, cost transparency and separation between factual status and advocacy. The institution earns deference by making deference less risky.
This bargain would lower the value of capture. If a board, staff office or faction can use dispute handling to shape address markets, actors will fight intensely to control it. If dispute handling is constrained, transparent and reviewable, control of the institution is less valuable as a private prize. Elections can then be more about governance than asset control, budgets more about services than war chests, and policy debates more about rules than disguised litigation.
AFRINIC cannot erase the history that made dispute resolution economically urgent. Alleged address-record abuse, the Cloud Innovation conflict, receivership, election controversies, continuing litigation and outside concern are now part of market memory. Institutions recover not by asking markets to forget, but by making the next dispute less damaging than the last. The question is whether the next contested resource, transfer, authority claim or court order produces a bounded status and credible forum, or whether it again becomes a test of survival.
The answer will affect more than reputation. It will influence the price of African IPv4 holdings, the willingness of operators to update records, the confidence of buyers and lenders, the bargaining power of small networks, the cost of leasing and transfers, and the ability of courts and coordination bodies to protect continuity without absorbing every registry fight. Dispute resolution is not a legal appendix to the registry model. It is the mechanism by which a scarce-resource ledger remains market infrastructure rather than a battlefield.
The neutral registry of the future is not the registry that never faces a serious dispute. It is the registry whose forum is strong enough that a serious dispute no longer threatens the credibility of the ledger itself.