Summary

  • APNIC has legitimate powers to keep the registry accurate, recover resources after defined failures, apply community policy, and protect the uniqueness of Internet number resources; those powers should be read as limited public infrastructure powers, not as open-ended corporate discretion.
  • The constitutional limits that matter most are mandate discipline, notice and reasons, proportionality, non-confiscation, transferability, portability, reviewability, capture resistance, and strict controls on emergency measures.
  • IPv4 scarcity has raised the stakes. A registry decision that once looked clerical can now affect market value, customer continuity, financing, merger execution, and a network's practical autonomy.
  • The bright line should be functional: APNIC may maintain the ledger and enforce resource policy, but should not use registry control as punishment for unrelated disputes, political pressure, institutional embarrassment, or member factional contests.
  • The best institutional guard is separation. Staff should administer records; policy forums should set general rules; elected bodies should supervise strategy; and contested deprivations of resource rights should be reasoned, reviewable, and proportionate.

The administrative act that carries constitutional weight

Start with an unglamorous act: an APNIC account fails to meet a contractual or operational condition, the registry sends notices, services may be suspended, and at the end of the sequence number resources can be recovered or marked as no longer held by that account. To the outside observer this is housekeeping. The registry is not raiding a data centre. It is not asking a carrier to tear out fibre. It is amending records in a database and withdrawing services around those records.

But that description is too small. In an address-scarce Internet, the registry ledger is not merely descriptive. It is the public memory that lets other networks, customers, banks, insurers, brokers, courts, and acquirers know who is recognised as the holder of a block. A resource holder can keep announcing a prefix for a time even after a registry dispute. The packet path is not the same thing as the legal and institutional path. Yet the practical value of the number resource depends on the surrounding recognition: reverse DNS, WHOIS and RDAP data, transfer records, route-registration habits, certification services, and the ability to prove clean title in a transaction.

That is why an RIR's everyday enforcement powers need constitutional thinking. The word constitutional is not used here to turn APNIC into a state. It is used in the more basic institutional sense: a body that controls a critical public ledger must be constrained by rules about purpose, method, remedy, and review. The more valuable the resource becomes, the more dangerous it is to treat an account closure, a transfer refusal, or a recovery decision as ordinary private administration.

APNIC is incorporated as a private organisation and is governed by corporate documents, membership agreements, by-laws, community policy, and operational guidelines. It also operates inside a global number-resource system that expects open policy development, regional service, and coordination with other registries. That mixture is useful because it lets engineers and network operators solve practical problems without waiting for governments. It is also fragile because private-law instruments can be asked to carry public-law burdens. A membership contract can say what happens when fees are not paid. It cannot by itself answer every question about proportionality when a registry act would destroy a scarce asset, strand customers, or favour one faction in a market dispute.

The Asia Pacific region makes the tension sharper. APNIC serves a region with mature carriers, island networks, hypergrowth platforms, universities, state operators, cloud providers, content networks, resellers, and small firms trying to buy or transfer addresses in a market where IPv4 is no longer abundant. The same registry act can be trivial for a well-capitalised carrier and fatal for a small operator with limited alternatives. Constitutional limits are a way of making the same ledger reliable for both.

What the APNIC mandate can and cannot bear

The first limit is mandate. APNIC's central job is to distribute, register, and administer Internet number resources for the region in a way that preserves uniqueness, accuracy, conservation, aggregation where relevant, and fair access under community-developed policies. It is not a general regulator of speech, competition, commercial morality, employment conduct, content, geopolitical alignment, or ordinary customer disputes. It may have to respond to court orders, fraud, sanctions, hijacking evidence, membership default, bankruptcy notices, and policy violations. But each response should be anchored to a registry purpose.

This distinction sounds obvious until a scarce resource becomes a lever. A body that controls valuable addresses will be tempted, or pressured, to use registry power to solve problems that are only loosely connected to the registry. A government may prefer a dissident network to lose standing. A dominant member may want a rival's transfer delayed. A complainant may ask for registry action because ordinary civil litigation is slow. A faction in a governance dispute may discover that the registry's control over account standing is a more effective weapon than argument.

Mandate discipline says no. APNIC can insist on accurate contacts because inaccurate contacts degrade the registry. It can require truthful transfer documents because false documents corrupt title. It can recover resources under defined policy where a holder has abandoned, misrepresented, or failed to maintain recognised status. It can apply the community's transfer rules because transferability without rules would turn the ledger into a clearinghouse for uncertainty. It can refuse to publish records known to be false because the database is a public reliance instrument. It cannot fairly convert every serious complaint into a resource sanction.

The point is not that non-registry harms never matter. Network abuse, insolvency, fraud, and sanctions can overlap with registry integrity. The point is that APNIC's coercive act must be justified in registry terms. Does the act protect uniqueness? Does it correct a false registration? Does it apply a published policy? Does it preserve the integrity of a transfer? Does it prevent the registry from being made a party to a fraud? If the answer is no, the dispute probably belongs elsewhere.

This is the most important constitutional line because every later safeguard depends on it. Due process without mandate discipline can produce well-administered overreach. A perfect hearing about an irrelevant ground is still overreach. Proportionality without mandate discipline can merely calibrate an improper sanction. Reviewability without mandate discipline can turn reviewers into managers of discretion rather than guardians of purpose. The first question in every severe registry act should be: what recognised registry purpose is being served?

Due process as a feature of registry accuracy

Due process is often treated as a favour granted to the affected party. In a registry, it is also a tool for accuracy. Notice, evidence, time to respond, reasoned decisions, and a neutral review path reduce the chance that the registry corrupts its own ledger by acting on bad information.

The minimum standard should vary with severity. Low-risk corrections can be fast. A stale contact can be marked, chased, and updated. A minor administrative failure can be cured. A suspected false transfer, a revocation of resources, a refusal to process a market transaction, or a disabling of critical registration services needs more. The affected holder should know the ground, the evidence, the policy or contractual provision invoked, the consequence, the cure period if any, and the path for review. When time is short, APNIC can act provisionally, but the provisional nature should be explicit and the post-action review should be real.

The due-process problem is not solved by saying that a member agreed to contractual terms. Consent to a membership agreement is meaningful, but it is not a complete answer when the registry is the only recognised regional source of the service. The resource holder cannot simply choose a competing Asia Pacific RIR. The registry is a natural monopoly by design. That does not make it a state, but it does mean ordinary private-law consent should be supplemented by public-infrastructure habits.

Reasons matter more than many institutions admit. A terse message that a request has failed does not help the community understand whether APNIC is applying policy consistently. A reasoned decision does. It identifies the rule, the facts, the judgment call, and the remedy. It can be redacted where necessary for security, privacy, or litigation. But the discipline of reasons constrains discretion. It also builds a record for future consistency.

Evidence matters because registry disputes often arise from documents that travel poorly across borders: company registers, mergers, insolvency proceedings, shareholder fights, court orders, asset-sale agreements, delegated authority letters, and legacy resource records. A registry serving dozens of legal systems must decide how much to believe and when to wait. That cannot be reduced to clerical form-filling. It requires an evidentiary standard appropriate to the consequence. A transfer delay may be justified by unresolved title evidence. A permanent recovery should require more.

Review matters because staff should not be the final judge of the most severe acts they initiate. Staff expertise is essential; it is not independence. APNIC's elected structures and community policy forums play different roles, but contested deprivation of resource standing needs a path that is separate from the original operational decision. The reviewer should be able to test mandate, evidence, proportionality, and remedy. A review that can only confirm that staff followed an internal form is not enough.

Proportionality in a scarce-address economy

Proportionality is the principle that the remedy should fit the registry harm. It is easy to accept in theory and hard to preserve in a market where IPv4 value is high. If a holder misses a payment, the registry has an interest in collecting fees and maintaining member discipline. If a holder submits false documents in a transfer, the registry has an interest in protecting the ledger. If a holder disappears, the registry has an interest in returning unused or abandoned resources to recognised status. These harms are not equal. The remedies should not be equal either.

A proportional system begins with graduated measures. Warning, clarification, contact validation, temporary suspension of non-essential services, transfer hold, restricted account functions, public status flags, and recovery are different instruments. They should not be collapsed into one blunt sanction. A serious registry integrity threat can justify a strong remedy. An ordinary fee dispute may justify service restriction and collection action before any step that threatens resource standing. A disputed corporate control issue may justify a temporary transfer hold without changing the underlying holder until the evidence clears.

The proportionality question has a time dimension. A temporary hold that is reasonable for two weeks may be unreasonable after two years. In a transfer market, delay has economic value. It can break a transaction, lower a price, favour a competitor, or prevent a network from financing expansion. A registry can damage a holder without ever issuing a final adverse decision simply by leaving a contested request unresolved. Constitutional limits therefore need not only final-remedy rules but timeliness duties.

Proportionality also requires attention to reliance. A network that has used a block for years, built customer assignments around it, documented routing arrangements, and kept registry contacts current has a different reliance claim from a shell entity trying to move disputed resources through opaque paperwork. That does not turn addresses into absolute property. It does mean the registry should distinguish between correction of a false or abandoned record and confiscation of a relied-upon operating asset.

The most dangerous move is to define every breach as an attack on registry integrity. If every failure can justify the most severe remedy, the limits disappear. Registry integrity is a real public good. It should not become a magic phrase that converts administrative convenience into confiscatory power.

Non-confiscation without pretending addresses are ordinary property

Internet number resources are not land. They are not intellectual property in the usual sense. RIR documents have long resisted the idea that allocations are unconditional property rights. That resistance is justified. Address space is globally coordinated; uniqueness depends on collective administration; and allocations were historically made under conservation and need rules rather than auction.

Yet the opposite proposition is also dangerous. Saying that addresses are not ordinary property does not mean they have no economic and reliance value. IPv4 scarcity has made the value visible. Transfers are common enough that market participants price blocks by size, reputation, region, documentation quality, and encumbrance risk. Lenders, acquirers, restructuring professionals, and operating networks treat recognised control of IPv4 as a valuable interest even if the legal label is carefully qualified.

A constitutional rule of non-confiscation can respect both truths. It need not say that a holder owns addresses absolutely. It can say that APNIC should not deprive a recognised holder of resource standing except under a published rule, for a registry purpose, with evidence, notice, proportional remedy, and review. It can say that recovery is legitimate where resources are abandoned, fraudulently obtained, no longer held by any recognised party, or subject to a defined policy that the community has adopted. It can also say that recovery is not legitimate merely because the holder is unpopular, commercially aggressive, politically inconvenient, or in dispute with a powerful member.

The distinction is especially important for historical resources and transfers. Legacy holdings often come with incomplete early records. Transfers can involve complex chains of corporate history. A registry must be able to ask hard questions. But uncertainty should be handled by targeted limits: require evidence, pause the transfer, mark the dispute, demand corporate documents, or seek legal clarity. Jumping from uncertainty to deprivation is where non-confiscation is violated.

Non-confiscation also protects APNIC itself. A registry that uses recovery sparingly and under clear rules is harder to accuse of opportunism. A registry that treats resource standing as revocable grace invites litigation, politics, and market discounting. The more participants fear discretionary confiscation, the more they will route around the registry through private warranties, complex indemnities, offshore vehicles, and pressure campaigns. That weakens the public ledger.

Transferability as a constitutional safety valve

Transferability is sometimes described as a market concession. It is more than that. In a post-exhaustion IPv4 world, transferability is a safety valve that lets resources move from lower-value to higher-value uses, lets new entrants buy access when free pools are gone, and lets corporate reorganisations preserve continuity. Without transfers, scarcity becomes administrative rationing. With transfers but no safeguards, scarcity becomes a title-risk market. The constitutional task is to allow movement while protecting the registry.

APNIC's transfer rules and public transfer log give the registry a legitimate role. It can require the parties to show authority, confirm that the recipient qualifies under policy, prevent double transfers, maintain history, and coordinate inter-registry movement. Those are ledger functions. But a transfer refusal should be tied to them. The registry should not block a transfer because it dislikes the price, the buyer, the seller's politics, or the strategic consequence, unless a published policy makes those considerations relevant.

The transfer context exposes how power can be exercised without formal confiscation. A transfer hold can immobilise an asset. A demand for repeated documentation can exhaust a small holder. A refusal to explain deficiencies can make cure impossible. A shifting interpretation of policy can change the economics after parties have signed. Each of these may be less dramatic than recovery, but the market consequence can be similar.

A sound constitutional practice would require transfer decisions to be predictable, documented, and appealable. Predictability does not mean automatic approval. It means the parties can know the rule before they act. Documentation does not mean publishing confidential sale contracts. It means the registry can explain the basis for approval, refusal, or delay. Appealability does not mean every disappointed buyer wins. It means a contested decision can be tested by someone other than the original decision-maker.

The registry also needs to preserve portability. A holder should not lose standing merely because it changes upstreams, uses addresses in multi-region services, or reorganises corporate structure, provided the use remains within policy and the registry record remains accurate. The Internet is no longer neatly regional in commercial design, even though RIRs remain regional institutions. Portability is the way the registry acknowledges that operational reality without abandoning regional accountability.

Separation between ledger maintenance and punishment

The cleanest constitutional line is the separation between ledger maintenance and punishment. Ledger maintenance asks: what must the registry do to keep records accurate, unique, current, and policy-compliant? Punishment asks: what hardship should be imposed on a party because of conduct the institution condemns? APNIC should do the first. It should be extremely wary of the second.

Some acts look punitive but are really ledger maintenance. If a party submits forged documents, rejecting the transfer is not punishment; it is refusal to corrupt the ledger. If resources were issued on a false premise, recovery can restore the integrity of the distribution system. If a holder cannot be identified, public status changes can warn others not to rely blindly on the record. If a court determines that a party lacks authority, the registry may need to follow the legal result.

Other acts are punishment dressed as maintenance. Cutting services to pressure payment beyond what is needed to collect fees, freezing unrelated resources because a member criticises the institution, withholding routine changes to gain leverage in a governance fight, or using registry status to settle a commercial dispute all move beyond the mandate. They may be tempting because they are effective. That is exactly why they need limits.

RPKI heightens the issue. Certification is an important routing-security service, but if registry status and certification are tightly bound, a registry sanction can have routing consequences beyond the database. The same is true of reverse DNS and routing-registry services. APNIC must preserve the ability to suspend services when the record is no longer reliable. But where the underlying resources remain in genuine dispute, the remedy should avoid unnecessary network disruption. The constitutional instinct should be to preserve operational continuity while the legal and registry questions are resolved.

This is not softness. It is infrastructure discipline. A registry earns legitimacy by being boring in the right way: consistent, evidence-based, slow to escalate, and precise about the harm it is fixing. The public interest is not served when an institution with a monopoly ledger behaves like an ordinary creditor or a factional actor.

Capture risk in a high-value registry

IPv4 scarcity changes governance incentives. When address blocks are cheap or administratively allocated, capture risk is lower. When a /16 or a larger historical holding can represent material value, the incentive to influence registry rules, elections, interpretations, and enforcement rises. APNIC's governance debates in the past several years have already shown that election design, nominee eligibility, voting rights, and member mobilisation are not decorative questions. They are part of the control surface around a valuable public ledger.

Capture need not look like a conspiracy. It can look like a membership class with more time to organise than ordinary operators. It can look like firms with large resource portfolios voting primarily to protect transfer value. It can look like incumbents favouring rules that burden new entrants. It can look like reformers overcorrecting and giving staff or elected bodies too much discretion to exclude inconvenient candidates. It can look like governments discovering that private registry structures are easier to influence than open technical forums.

The constitutional answer is not to distrust every participant. The address community is full of people who understand that the registry's value depends on neutral administration. But good motives are not a control system. APNIC needs governance arrangements that make capture hard: clear by-laws, transparent elections, conflict rules, public rationales for major reforms, independent election administration where appropriate, predictable voting rights, and open policy forums where the cost of manipulation is visible.

Capture risk also argues for narrow enforcement discretion. If a captured or partially captured institution can deprive holders of resource standing through broad clauses, the market will notice. Scarce assets will carry political risk discounts. Smaller networks will fear rule changes they cannot influence. Cross-border holders will ask whether regional governance can be trusted. The registry's neutrality is therefore not a moral luxury; it is an economic asset.

The same logic applies to emergency powers. A registry needs the ability to act quickly when uniqueness, fraud, hijacking, sanctions compliance, court orders, or security incidents create immediate risk. But emergency power is the classic capture doorway. Measures adopted under pressure tend to persist. Temporary suspensions become practical outcomes. Confidential evidence becomes a habit. The constitutional rule should be narrow trigger, written reasons, least-disruptive remedy, short duration, swift review, and public reporting once sensitive details can be protected.

Reviewability and the value of reasons

Reviewability is the difference between discretion and authority. A decision is not reviewable merely because the affected party can send another email. Reviewability means there is a defined body or path with the authority to examine the evidence, test the rule, consider proportionality, and alter the outcome. For the most serious registry acts, this is essential.

There are several layers. Operational review can catch mistakes quickly. Senior staff review can test consistency. Elected oversight can ask whether staff are using powers in a way that fits institutional purpose. Community policy can change general rules if outcomes reveal a design problem. Courts can address contractual, corporate, or property-like disputes when private law is implicated. None of these layers is perfect. Together they reduce the risk that one office becomes the unreviewed owner of the ledger.

APNIC should prefer review methods that preserve technical continuity. Litigation is sometimes unavoidable, but it is often a poor first instrument. It is slow, expensive, and difficult across jurisdictions. A registry-specific review path can decide whether a transfer hold is justified, whether evidence is sufficient, whether a cure period should be extended, or whether a recovery decision is premature. That kind of review does not replace courts. It can prevent unnecessary escalation.

Public reasons also educate the market. If APNIC publishes anonymised or redacted summaries of severe resource actions, participants learn what conduct actually triggers strong remedies. If it publishes nothing, rumours become the law. A market ruled by rumours is inefficient and unfair. It favours insiders who know how the registry tends to think and penalises smaller holders that cannot afford specialist counsel.

The challenge is confidentiality. Registry disputes often involve private contracts, security incidents, personal data, and legal claims. Public reasons can be calibrated. APNIC can explain the category of issue, policy basis, remedy, and review result without exposing sensitive details. The constitutional point is not total transparency. It is enough transparency to make power intelligible.

How APNIC should think about emergencies

Emergency powers are necessary because the Internet does not wait politely. If a resource record is being used to facilitate obvious fraud, if a court order binds the registry, if a sanctions rule creates a legal prohibition, or if a hijacking claim requires immediate status protection, APNIC may need to act before a full record is assembled. The constitutional question is how to keep emergency action from becoming ordinary governance.

First, the trigger should be narrow. "Urgent" should mean a concrete risk to registry integrity, legal compliance, routing security, or public reliance, not reputational pressure or political discomfort. Second, the remedy should be the least disruptive measure that addresses the risk. A temporary transfer lock may be enough where title is disputed. A contact validation flag may be enough where reachability is uncertain. Full recovery should be rare before review unless the record is plainly false or the holder has abandoned any recognised claim.

Third, emergency acts should expire unless renewed through a reasoned decision. Time limits force the institution to build a record. Fourth, the affected party should receive timely notice unless notice would defeat the remedy or violate law. Fifth, independent review should be available after the fact. Sixth, aggregate reporting should show the community how often emergency powers are used and in what categories.

Emergency reporting has another benefit: it lets the policy community see whether existing rules are too weak. If staff repeatedly need emergency discretion to address the same problem, the answer may be a policy change rather than more discretion. A healthy registry converts repeated exceptional cases into general rules. An unhealthy one lets exceptions accumulate until they become the real constitution.

External law and registry restraint

APNIC does not operate in a legal vacuum. It may receive court orders, insolvency notices, sanctions questions, police inquiries, civil claims, and competing corporate-control documents from many jurisdictions. The hard part is not admitting that law matters. It plainly does. The hard part is deciding what a regional registry should do when law and registry standing intersect.

The safe constitutional instinct is restraint. APNIC should obey binding legal obligations, but it should avoid converting every legal allegation into registry action. A lawsuit between shareholders does not automatically mean the resource record is false. A commercial claim by a creditor does not automatically create a right to seize addresses. A government letter is not always a court order. A foreign proceeding may be relevant evidence without being conclusive in the jurisdiction where the account holder is organised. A registry serving the Asia Pacific region needs a method for reading legal materials without becoming a general court for the region's business disputes.

That method should separate recognition from enforcement. Recognition asks who APNIC can safely treat as the authorised resource holder or account controller for registry purposes. Enforcement asks whether a private claimant can collect a debt, enforce a security interest, punish misconduct, or obtain damages. The first question often belongs to APNIC because the ledger must name someone. The second usually belongs to courts and arbitral bodies. Trouble begins when a registry remedy designed for recognition is used to enforce a private claim.

Sanctions and public-law prohibitions are more difficult because APNIC may itself face legal exposure. Even then, the constitutional habit should be narrow tailoring. If a legal rule prohibits service to a named party, the registry may have to restrict service. It should still preserve accurate historical records, document the basis of the restriction, avoid wider collateral harm where lawful, and provide a path to correct mistaken identity. A sanctions response should not become an excuse to alter ownership-like standing beyond what the law requires.

Insolvency presents another test. Address resources may be part of a restructuring plan, asset sale, or disputed estate. The registry must protect itself against fraudulent transfers and unauthorised signatories. But it should also recognise that insolvency law often tries to preserve going-concern value. A transfer delay that destroys a sale can reduce creditor recoveries and disrupt customers. APNIC's role is not to maximise the estate. It is to maintain a clean registry while respecting credible legal authority. That calls for temporary holds, evidence requests, and reasoned decisions, not automatic deprivation.

The same caution applies to abuse complaints. Network abuse is real. Spam, botnets, malware distribution, routing hijacks, and fraud impose costs on others. But APNIC is not a content moderator or universal network police body. It can require accurate contacts and may act when registry services are themselves being misused. It should be wary of resource sanctions based solely on allegations that are better handled by operators, law enforcement, courts, or security coordination channels. If resource standing becomes the universal abuse remedy, the registry will be drawn into cases it is not designed to adjudicate.

External law therefore reinforces the mandate limit rather than weakening it. APNIC should be responsive to lawful compulsion and credible evidence. It should not be pliable. A private registry with a monopoly ledger is most legitimate when it can tell powerful outsiders that its tools are limited to registry purposes and that broader punishment must come from the institutions authorised to impose it.

What resource holders owe in return

Constitutional limits are not a licence for resource holders to treat the registry as a passive title vault. APNIC's restraint depends on reciprocal duties. Holders should keep contacts accurate, pay legitimate fees, maintain authority records, answer validation requests, document corporate changes, use transfer channels honestly, and avoid presenting forged or misleading materials. A holder that wants the protection of due process should not frustrate the facts that make due process possible.

This reciprocity matters because bad-faith holders exploit every safeguard. A party that disappears for years and then complains about recovery after repeated notices weakens the case for patience. A transferor that withholds corporate documents and then denounces delay weakens the case for speed. A buyer that structures a deal to evade eligibility rules weakens the case for transferability. A member that uses governance rights only to defend speculative address value weakens the case for community trust.

The constitutional answer is not to abandon safeguards when holders behave badly. It is to let bad faith matter as evidence and remedy. A party that ignores notices may receive shorter cure periods. A party that submits contradictory documents may face a transfer hold. A party that lies can be denied relief. A party that repeatedly abuses reserve or transfer rules can be restricted under published policy. Due process does not require institutional naivety.

What it does require is that the registry name the misconduct and connect it to the remedy. If the problem is false documentation, the remedy should protect the document-dependent act. If the problem is non-payment, the remedy should collect fees and maintain member discipline without unnecessary destruction of resource standing. If the problem is abandoned contact, the remedy should restore reachability or mark uncertainty before recovery. If the problem is legal uncertainty, the remedy should preserve the status quo long enough for authority to be established.

This reciprocal frame is important for community politics. Holders sometimes describe any registry constraint as confiscation. That is wrong. APNIC is entitled to administer a public ledger under public rules. It is entitled to say no. It is entitled to demand evidence. It is entitled to protect other users from false records. The constitutional claim is narrower and stronger: because APNIC has those powers, it should use them for registry ends, with reasons and review, and with remedies that fit the harm.

The APNIC constitution is a market institution

The phrase "constitutional limits" can sound legalistic, but in the RIR setting it is also an economic argument. Markets in IPv4 addresses depend on confidence that the registry will recognise clean transactions, reject bad ones, and refrain from opportunistic intervention. Network operators depend on confidence that routine administration will not become a threat to continuity. The community depends on confidence that scarce resources are governed by rules, not factions.

APNIC's limits therefore support its power. A registry that is tightly constrained in purpose is more trusted when it must act. A registry that gives reasons is more credible when it refuses a transfer. A registry that uses proportionate remedies is more likely to be obeyed. A registry that separates maintenance from punishment is less likely to be drawn into commercial or political disputes. A registry that invites review is less likely to make uncorrected errors.

This is not a call for paralysis. APNIC should recover abandoned resources. It should reject fraudulent transfers. It should enforce published policy. It should protect the accuracy of contacts and resource records. It should maintain robust election rules and defend the institution against capture. It should comply with law. It should act quickly in genuine emergencies. The question is not whether APNIC has power. It is whether that power is bounded by the kind of discipline appropriate to a monopoly public ledger.

The constitutional standard can be expressed simply. Severe action against resource standing should be lawful under APNIC's rules, tied to a registry purpose, based on evidence, preceded by notice where possible, proportionate to the harm, protective of operational continuity where feasible, open to review, and separated from unrelated punishment. Transfers should be blocked only for defined policy, authority, fraud, sanctions, title, or registry-integrity reasons. Emergency measures should be narrow, temporary, reasoned, and reviewed.

There is a practical reporting corollary. APNIC need not publish confidential case files, but it should be able to tell the community how often severe account restrictions, transfer holds, recoveries, emergency measures, and successful reviews occur in broad categories. Aggregate numbers would not settle every dispute, yet they would reveal whether strong powers are rare exceptions or ordinary management tools. In a scarcity economy, that distinction is itself governance information.

There is also a design corollary for ordinary tools. Account portals, transfer forms, status labels, and service notices should be built to preserve the same distinctions the policy claims to honour. A form that offers only approval or refusal can push staff toward blunt outcomes. A status label that cannot distinguish temporary dispute, missing evidence, legal hold, non-payment, and recovery can mislead the market. Constitutional restraint is partly a matter of institutional language, but it is also a matter of operational vocabulary.

The future will make this more important, not less. IPv6 adoption reduces the technical need for IPv4 over time, but it does not erase the commercial value of existing IPv4 holdings in the medium term. The transfer market will remain. Corporate reorganisations will remain. Government interest in Internet infrastructure will grow. Election contests will sharpen whenever registry decisions affect scarce assets. APNIC's legitimacy will rest not on declaring itself neutral but on making neutrality enforceable.

Sources and Further Reading