Summary

  • A court order must be obeyed, yet an Internet number registry is not an ordinary office: disabling its shared systems can transmit a dispute to members and customers who were never before the court.
  • Contempt analysis should identify the exact command, the person able to perform it, the evidence of knowledge and ability, and the period of non-compliance before coercion is selected.
  • The strongest enforcement design targets responsible officers, money, reporting duties and disputed transactions while ring-fencing registration data, routing-security administration, reverse DNS and security response.
  • AFRINIC's litigation history makes continuity planning a present governance duty, not an argument for immunity; published judicial and institutional materials should be tested as evidence rather than repeated as advocacy.
  • A durable transition, whether through repaired RIR governance or a future Number Resource Society arrangement, requires a court-ready map of authority, custody, substitutes and essential services.

The wrong choice is obedience or continuity

When a court directs an institution to act, institutional importance does not create a licence to ignore the order. The rule of law would become optional precisely where concentrated power makes compliance most important. An Internet number registry can affect access to registration records, resource certification, reverse DNS, transfers and the evidence by which operators demonstrate administrative control. That makes judicial enforceability more necessary, not less.

It also makes the design of enforcement unusually consequential. A conventional company may be pressured through its bank accounts, premises or officers without affecting an entire regional coordination function. A regional registry occupies a shared administrative position. If coercion disables the database, credentials, nameservers or staff needed for routine maintenance, the cost can move from the litigants to thousands of networks. Those networks may know little about the underlying case and possess no means to cure the disobedience.

The resulting governance question is not whether continuity outranks the court. It is how to secure compliance without converting third parties into enforcement collateral. Court authority and service continuity are complementary objectives when responsibility is accurately located. A judge can direct the individual who controls an act, require a verified report, restrain a disputed transaction, appoint a substitute for a defined task or impose a financial consequence while preserving the functions that keep unrelated resource records usable.

The alternative is a false binary. One side invokes the Internet to resist ordinary accountability. The other treats every institutional asset as an interchangeable pressure point. Both positions conceal the practical work: identifying the command, the responsible actor, the minimum coercion likely to work and the services that must remain available while the dispute is resolved.

Contempt begins with the command, not the institution's reputation

A serious contempt inquiry starts with the order itself. What exactly had to be done or avoided? When did the obligation begin? On whom was it binding? Was the command conditional? Did it require a board resolution, an officer's signature, access to a system, delivery of documents or restraint from a particular transaction? These details determine whether non-compliance is established and who can cure it.

Public debate often begins elsewhere. One party describes a history of obstruction; another describes institutional persecution. Those narratives may contain relevant facts, but they are not substitutes for construing the operative words. A court should not punish a general atmosphere. Nor should an institution evade a precise command by answering a broader accusation that the court did not make.

Clarity has a continuity benefit. If the required act is narrow, the enforcement response can remain narrow. An order to provide a report does not normally require interference with resource records. A direction governing an election does not by itself demand suspension of reverse DNS. A restraint on transferring a disputed asset need not disable members' access to routine services. The more carefully the command is separated from surrounding conflict, the easier it becomes to protect unaffected functions.

This is also why parties should seek clarification promptly when an order is genuinely ambiguous. Silent non-performance followed by a broad claim of uncertainty is weak governance. A written request identifying the doubtful words, the proposed compliant act and the operational consequences gives the court something concrete to decide. It creates a record of good-faith effort without allowing ambiguity to become indefinite delay.

Knowledge, ability and control must be mapped separately

Non-compliance by a legal entity is performed through people. The company has no hands, password or independent intention. Directors, receivers, executives, employees, service providers and custodians hold different parts of the capacity to comply. A continuity-aware court therefore needs a control map rather than a generic instruction to "the registry."

Knowledge is the first element. Which person received the order or was formally notified? Ability is the second. Could that person lawfully and practically perform the required act? Control is the third. Did another officer hold the credentials, records, bank authority or corporate power needed? These questions can point to different individuals. A director may know of the order but lack access to a system controlled by a receiver. A technician may possess access but lack authority to approve the transaction. A contractor may maintain infrastructure but be unable to change corporate records.

Mapping these roles prevents two errors. The first is punishing an individual for an act outside that person's capacity. The second is allowing responsibility to disappear into organisational complexity. If one actor holds legal authority and another holds technical custody, the order can require coordinated performance and verified handover. If access has been withheld, the record should identify by whom, since when and under what asserted authority.

The map should be time-specific. Control can change during litigation. Board composition, receivership powers, vendor access and staff employment may shift. An order directed to yesterday's officeholder may need variation rather than theatrical enforcement. But a change in office should not erase the institution's duty. It should trigger a documented transfer of knowledge, authority and credentials so the obligation follows the function.

A registry is a layered service, not one switch

The phrase "registry operations" is too broad for either judicial protection or institutional defence. Number administration contains separable functions. Registration data records allocations and assignments. Account systems identify authorised contacts. Resource Public Key Infrastructure supports route-origin authorisations. Reverse DNS delegation supports address-to-name functions. Transfer handling changes administrative control. Billing, membership voting, new-resource requests, public communication and abuse contacts are additional layers.

Each layer has a different relationship to a dispute. A contested election may concern who can appoint officers without making existing routing-security entities inaccurate. A contractual claim may concern money without requiring changes to reverse DNS. A dispute over one member's resources may justify preserving a specific record while leaving other members untouched. Treating all services as a single asset invites excessive coercion and excessive claims of fragility.

Layering also exposes dependencies. A service may appear separable but share credentials, staff or infrastructure with another. The court needs to know whether restraining payments would prevent nameserver renewal, whether replacing an officer would interrupt certificate signing, or whether a vendor can continue read-only operation while control is contested. These are evidentiary questions, not reasons to accept whichever side uses the language of stability most confidently.

A useful continuity schedule lists each function, its custodian, the minimum staffing and funding required, its recovery time, the consequence of interruption and the lawful substitute able to maintain it. That schedule gives a judge alternatives. Instead of choosing between unrestricted management and institutional paralysis, the court can preserve specified services while directing compliance elsewhere.

The official record is evidence, not a script

AFRINIC publishes corporate documents, policy materials, communiques and copies of court decisions. Mauritius publishes company law and public parliamentary records. ICANN and the RFC Editor publish descriptions of the wider registry system. These materials are essential, but their institutional origin matters.

A judgment is evidence of what the court decided and, subject to its terms, findings made on the record. It is not proof of every allegation recited from the parties. A corporate communique is evidence of what the institution publicly said; it is not an independent finding that the account is complete. A parliamentary statement shows official concern and political understanding; it does not decide private rights. A technical architecture document describes coordination goals; it does not resolve which coercive remedy Mauritian law supplies in a particular case.

This discipline is especially important in a prolonged institutional dispute. Repetition can make a contested claim appear settled. Labels such as obstruction, capture, emergency or continuity may migrate from pleadings into public commentary without the underlying evidence travelling with them. An evidence-led analysis asks who made the statement, under what authority, on which date, for what purpose and whether a court adopted it.

The same caution applies to silence. Missing public documents do not prove that no report, payment, instruction or attempt at compliance existed. They define an uncertainty. Where the issue is contempt, the decisive record should include the sealed order, proof of notice, correspondence about performance, affidavits from the persons with control and any subsequent clarification. Governance commentary should not manufacture certainty where that record is incomplete.

Judicial authority loses value when compliance is symbolic

An order that exists only on paper does not protect a member, an election, a company or the authority of the court. Delay can determine outcomes. A meeting date passes. Funds move. Staff leave. Evidence becomes harder to preserve. A supposedly temporary governance arrangement becomes normal. Compliance after the practical value of the order has expired may be formally neat and substantively empty.

That is why enforcement must be capable of changing behaviour in time. The responsible actor should know the deadline, the evidence required to demonstrate performance and the consequence of continued refusal. Progress reports should distinguish completed acts from planned acts. If a dependency prevents performance, the report should identify the dependency and request a direction before the deadline rather than after it.

Time also changes the appropriate coercive tool. Early in non-compliance, clarification and a short compliance timetable may be sufficient. Repeated missed deadlines after clear notice may justify personal attendance, sworn disclosure, daily financial pressure or appointment of a substitute. The escalation should respond to demonstrated failure of the lower measure. It should not begin with infrastructure disruption merely because infrastructure is the institution's most visible asset.

The objective is effective obedience. Punishment can vindicate authority, but coercive contempt is most useful when the subject can end the pressure by complying. A remedy that destroys the capacity to perform the order defeats itself. If accounts are frozen so completely that staff cannot retrieve documents, maintain services or convene the directed meeting, the sanction may create a new explanation for failure. Precision keeps enforcement both credible and productive.

Personal responsibility should not be hidden inside the company

Corporate form can blur accountability. Decisions are described as actions of "AFRINIC" even when a named officeholder approved, delayed or prevented them. The institution then bears reputational and financial cost while the individual decision-maker remains indistinct. That is poor deterrence and poor continuity design.

Where the law and the order permit, enforcement should follow actual responsibility. A person who was bound, knew the command, had capacity to comply and chose not to do so is a more coherent target than a shared service used by innocent members. Personal attendance can test explanations. A sworn statement can expose who holds control. A personal costs order or other lawful consequence can prevent the wrongdoer from externalising the price to the registry's operating budget.

This does not mean presuming misconduct by directors, staff or a receiver. Office alone is not proof. Collective decisions may require a quorum; legal advice may identify conflicting duties; credentials may genuinely be unavailable. The court should hear those facts. The point is to investigate the human chain rather than letting the legal entity function as both shield and scapegoat.

Personal targeting also protects successors. A newly appointed board or officer should not inherit unbounded punishment for an earlier actor's defiance if it promptly discloses the position and complies. Institutional obligations continue, but culpability and remedial capacity can differ. Separating them gives new leadership an incentive to restore obedience quickly instead of defending every prior decision as a condition of organisational survival.

Money is a lever, but the source of money matters

Financial sanctions are attractive because they avoid direct interference with technical systems. Yet a fine paid from membership fees may still transfer the cost to parties who did not disobey. It can reduce funds available for security, staffing, audits or service resilience. The court therefore needs to understand incidence, not merely amount.

A financial order should identify whether payment falls on the company, a responsible officeholder, an insurer or a protected fund. It should consider whether the amount is coercive without consuming the working capital required for essential service. Where personal responsibility is proved and law allows, personal incidence may be more accurate. Where the company is properly liable, ring-fencing the minimum operating budget can preserve continuity without making the institution judgment-proof.

Escalating daily sums can be effective when the required act is clear and immediately possible. They are less useful when performance depends on a disputed appointment, inaccessible data or third-party consent. In that situation, the sanction should be paired with disclosure and a route for resolving the obstacle. Otherwise money accumulates while the command remains undone.

Freeze the disputed transaction before the shared service

Many institutional conflicts concern a specific act: transferring assets, convening an election, recognising directors, paying a disputed claim, changing a record or disposing of property. The first continuity question is whether that act can be restrained or performed without touching unrelated services.

A transaction-specific order has informational advantages. It states what cannot change and leaves the ordinary baseline intact. Staff and vendors can continue approved functions. Members can understand that the restriction is bounded. The court can later lift or vary it without reconstructing an entire operating environment. The parties also have less opportunity to portray every routine decision as either prohibited or authorised.

Specificity requires careful drafting. A prohibition on "dealing with resources" could be interpreted so broadly that even maintenance becomes risky. A better order can distinguish reassignment, transfer or deletion from routine accuracy updates, security response and preservation. If a disputed record must not change, a neutral custodian can retain an immutable copy while authorised technical maintenance continues under logged controls.

The same principle applies to funds. A restraint over a contested payment does not necessarily require freezing payroll, hosting or certificate infrastructure. If dissipation is feared, designated accounts and dual authorisation may protect the claim. The court should receive evidence about expected operating expenses and the risk of concealment. Continuity is preserved through supervised granularity, not trust alone.

Verified reporting is an enforcement instrument

Courts often face an information problem before a compliance problem can be solved. The institution knows who holds credentials, what actions occurred, which vendors are critical and what remains possible. The opposing party may know only the visible outcome. A verified report can convert that asymmetry into an enforceable record.

The report should answer operational questions, not offer a narrative of institutional virtue. It can list each required act, the responsible person, completion evidence, blockers, next deadline and continuity effect. It can identify relevant accounts, contracts, repositories and access holders without publishing sensitive details. Supporting records can be provided confidentially where security or personal data requires it.

Verification raises the cost of strategic ambiguity. A statement that compliance is "under way" becomes a series of testable assertions. If a password is unavailable, the report names the custodian and recovery request. If a board resolution is required, it identifies the meeting and votes. If a vendor refuses an instruction, the correspondence is produced. The court can then direct the real bottleneck.

Reporting should not become delay by documentation. The timetable must be proportionate to urgency, and the court can require immediate disclosure of decisive facts followed by a fuller account. Nor should sensitive infrastructure data enter an unrestricted public file. A confidentiality protocol can protect attack surfaces while preserving adversarial testing by authorised counsel or an expert.

A substitute can perform the act without taking the institution

When the responsible actor will not comply, appointment of a substitute for a defined task may be more effective than broad receivership or disabling sanctions. The substitute might sign a document, deliver records, convene a meeting, preserve evidence or implement a specific database instruction under court authority. The exact availability and form depend on applicable law, but the design principle is general.

Task-specific substitution preserves organisational boundaries. It does not assume that the court must manage the registry indefinitely. It solves the blocked act and returns ordinary authority once complete. The substitute's mandate, access, reporting duty, liability and end point should be explicit. Vendors and staff need to know which instructions are valid and which remain with ordinary management.

Technical acts require particular care. A substitute should not receive unrestricted credentials if a narrower signed instruction or supervised session will do. Changes should be logged, peer-reviewed and reversible where possible. An independent expert can confirm that the act matches the order and does not alter unrelated records. These controls protect both the court and the registry from accidental overreach.

Substitution also changes incentives. An officeholder cannot preserve a veto by refusing to act, yet the institution is not punished through general paralysis. If the blocked actor later cooperates, the court can restore the normal chain. The remedy is therefore coercive in effect but conservative in institutional scope.

Receivership must have a service boundary

A receiver can provide lawful custody, reporting and authority when ordinary governance cannot function. AFRINIC's published litigation record makes the question practical rather than hypothetical. But "receiver" is a legal role, not a complete operating model for an Internet registry.

The appointment instrument should distinguish company preservation, litigation compliance, elections, financial control and technical registry operations. A receiver may be expert in insolvency and still need specialist support for routing-security systems, database integrity, cryptographic ceremonies and member authentication. Delegation of technical work should be explicit, supervised and auditable. Silence invites either dangerous improvisation or overreliance on incumbent staff whose authority may be contested.

The receiver also needs a continuity budget, vendor map, incident plan and conflict register. Which contracts may be renewed without further approval? Which actions require court directions? Can the receiver alter resource records, or only preserve them? Who reviews emergency security changes? How are member complaints escalated? These questions should be answered before crisis forces an answer.

Receivership must include an exit path. Milestones can cover restored accounts, verified records, audited finances, valid governance appointments and transfer of credentials. Without them, extraordinary administration may persist because no one has defined completion. Long duration can itself change institutional incentives: staff adapt to uncertain authority, members disengage and vendors price risk. A service boundary and termination plan preserve the receiver's legitimacy.

Essential services should be ring-fenced by function

Continuity protection should attach to functions, not to every preference of management. The core ring may include authoritative registration data, security response, reverse DNS delegation, routing-security maintenance, credential recovery, backups and minimum member support needed to prevent harm. New discretionary programmes, travel, public campaigning or contested strategic projects do not acquire the same protection merely because the institution operates them.

Functional ring-fencing requires minimum-service definitions. How quickly must a security incident be addressed? Which changes are permitted during a governance dispute? What authentication is required? How often are backups tested? Which staff and vendors are essential? The answers should be realistic enough to operate and narrow enough not to immunise ordinary spending.

The ring should also preserve evidence. Logs, database snapshots, signing records and access histories may be both operationally vital and relevant to litigation. Retention controls can prevent deletion while allowing ordinary use. A neutral copy held under sealed conditions may resolve later disputes about what changed and when.

Members should receive a plain explanation of protected services and temporary restrictions. They do not need confidential pleadings to plan their networks. They need to know whether maintenance, certificates, transfers, billing and support remain available, how to report urgent problems and what fallback applies. Accurate service communication reduces rumours that can cause more commercial damage than the formal order.

Continuity is not immunity from evidence preservation

Institutions sometimes invoke security to resist disclosure. Some restrictions are legitimate: unrestricted publication of credentials, signing details or vulnerability information can cause harm. But security cannot become a category into which every inconvenient record disappears.

A court can separate preservation, access and publication. Relevant data can be imaged and hashed, held by a neutral expert, reviewed under confidentiality and disclosed only to authorised persons. Privilege, personal data and security sensitivity can be assessed document by document or category by category. The important point is that the information survives while those questions are decided.

Evidence preservation also serves continuity. Logs reveal unauthorised access. Configuration records support recovery. Vendor contracts identify renewal and step-in rights. Staff rosters show who can perform critical tasks. Destruction or fragmentation of this evidence weakens both adjudication and operations.

The preservation order should be technically literate. A direction to retain "all data" may be impossible or obscure the important systems. The schedule can identify mail, board records, financial ledgers, ticket systems, source repositories, identity systems, resource databases, RPKI records and backups. It can specify date ranges, custodians and retention methods. Routine deletion should pause only where relevant, because unbounded preservation imposes cost and privacy risk.

The burden of continuity evidence belongs to the institution that holds it

A registry seeking special protection for essential services should provide evidence of the dependency. It should not merely assert that any sanction threatens the Internet. The institution knows its systems, staffing, vendors, cash needs and recovery arrangements. It is best placed to show which functions are at risk and what narrower orders are available.

That burden does not require disclosing attack-enabling details publicly. A sealed continuity statement can identify architecture, minimum funds, privileged accounts and recovery procedures. An independent expert can test whether the claims are technically credible. The opposing party should receive enough information to challenge overbreadth without receiving secrets irrelevant to the dispute.

The court should distinguish unavoidable risk from remediable governance weakness. If one person's password can stop a regional service, that is not a permanent reason to protect the person from enforcement. It is evidence that credential redundancy must be repaired. If no tested backup exists, the remedy may need a short stabilisation period, but the court can require a backup and recovery test rather than accept indefinite vulnerability.

Continuity claims are strongest when they come with alternatives. An institution that says "do not freeze this account; designate this smaller account, preserve these vendor payments and require dual approval" is assisting the court. An institution that offers only immunity is asking dependent networks to subsidise unaccountable control.

Emergency orders need rapid review

Courts sometimes must act before a full record exists. Assets may move, credentials may change or a meeting may create irreversible consequences. An interim order can preserve the position. Because emergency relief is based on compressed evidence, it should also include a short review clock and clear return date.

The first order should identify what danger is being contained, which services remain authorised and what evidence the parties must provide. If a broad restraint is unavoidable for hours or days, the review should narrow it once the system map is available. Emergency breadth should not become permanent through inertia.

Registry officers must maintain an emergency contact and an accurate list of critical functions so they can respond immediately. Claiming that no one can explain the impact until weeks later is itself a resilience failure. The institution should be able to show the court what can safely continue under read-only, dual-control or supervised conditions.

Rapid review also protects judicial authority. An order that unexpectedly disrupts services can create pressure for abrupt reversal and public blame. A designed review point allows correction without implying that the original protective purpose was illegitimate. Courts manage complex assets more credibly when they treat operational evidence as dynamic.

Compliance should be proved by outcome, not ceremony

An institution may produce minutes, resolutions and letters while the required result remains undone. Conversely, a technical result may occur without the corporate authorisation the order demanded. Proof of compliance should match the command and include both authority and effect where both matter.

If the order requires an election, evidence may include valid notice, voter eligibility, nomination handling, ballot custody, counting, results and lawful assumption of office. If it requires restoration of access, a screenshot is weak; authenticated access, changed credentials, vendor recognition and a test transaction are stronger. If it requires delivery of records, the receiving party should be able to confirm completeness or identify gaps.

Independent verification is valuable where trust has collapsed. A technical expert, auditor or officer of the court can confirm completion without deciding the whole dispute. Verification should be scoped and time-limited. It is not an invitation to create a parallel permanent management body.

Outcome proof also reveals partial compliance. Some steps may be complete while another remains blocked. The court can then target the unresolved element instead of treating the institution as wholly obedient or wholly defiant. Granularity improves both fairness and coercive effectiveness.

Appeals do not silently suspend obligations

An institution entitled to appeal should use that right. It should also seek a stay when performance before appeal would cause irreversible harm. Filing an appeal does not automatically answer whether the order remains binding; that depends on applicable law and the court's directions. Governance communication must not blur the distinction.

The compliance record should state whether a stay was requested, what relief was granted and which obligations continue. Staff and vendors should receive the operative position, not assumptions. If only part of an order is stayed, the remaining acts should proceed. If the court refuses a stay, the institution should comply while preserving its legal arguments unless another lawful direction applies.

Continuity evidence is relevant to a stay application, but it must be specific. The applicant should show the likely service effect, why later restoration is inadequate and what safeguards can protect the opposing party. General claims about regional importance are less persuasive than a function-by-function plan.

Institutional redesign should follow the failure mechanism

After a crisis, organisations often promise broad reform. Useful reform begins with the actual mechanism of failure. Was the order unclear? Did one officer hold all credentials? Could the board not form a quorum? Did vendors reject lawful successors? Were funds commingled? Did members lack information? Each mechanism suggests a different repair.

Credential concentration calls for role-based access, multiple custodians and tested succession. Governance deadlock calls for vacancy rules and emergency appointment procedures. Unclear authority calls for a current mandate register. Financial opacity calls for segregated budgets and reporting. Weak court responsiveness calls for service addresses, legal escalation and a standing order-response protocol.

The reforms should be tested through exercises. A paper continuity plan may fail when a vendor asks for an unanticipated document or a backup cannot be restored. Simulated loss of an officer, temporary account restraint and disputed instruction can reveal the real dependencies. Results should be reported to members at an appropriate level.

Reform also needs ownership and deadlines. "Improve resilience" is not a control. "Rotate signing custody among three authorised roles, test recovery twice yearly and report exceptions to the audit committee" is. The institution proves learning when the former failure can no longer recur through the same path.

Number Resource Society can learn without claiming succession by slogan

Number Resource Society is sometimes presented as a future direction for placing operators and resource holders closer to the centre of number governance. That direction is useful only if it solves the enforcement and continuity problems revealed by existing arrangements. A new name does not distribute authority, preserve evidence or answer a court order.

A credible design would include a mandate ledger showing who can instruct each essential function, a continuity charter separating protected services from discretionary programmes, and a court-response rule that names responsible officers and substitutes. It would require independent custody arrangements, conflict disclosure, member-visible legal expenditure and rapid review of emergency restraints.

The Society model could also make portability and institutional substitution less catastrophic. If data standards, credentials, escrow and service interfaces are designed for lawful transfer, accountability no longer depends on preserving one management group at all costs. That does not imply casual replacement of a regional registry. It means the system should survive governance failure without holding operators hostage.

Positive direction must remain conditional. The Society would need legitimate regional support, technical competence, sustainable funding, legal recognition and disciplined boundaries. It should not use dissatisfaction with AFRINIC as proof of its own authority. Its case should rest on auditable design and demonstrated service capacity.

A court-ready registry is a better registry

Preparing for judicial enforcement may sound adversarial, but it is ordinary institutional maturity. Contracts are written so successors can perform them. Financial systems preserve audit trails. Critical infrastructure has disaster recovery. A registry should likewise know how to obey an urgent order without losing control of essential services.

Court readiness includes a verified legal service address, current officeholder register, order triage, privilege protocol, evidence-preservation schedule, system control map, vendor recognition pack, continuity budget and substitute-authority mechanism. It includes training for executives and technical staff so neither improvises under pressure. It includes a board rule that disputed legal strategy cannot interrupt protected operations without explicit authority.

These controls deter both disobedience and opportunistic litigation. A claimant knows that an order can be performed precisely. Management knows that institutional fragility cannot be used as a shield. Members know that their services will not be switched off to increase bargaining pressure. Courts receive better options than paralysis or indulgence.

The cost is modest compared with prolonged crisis. Most of the required information should already exist for security, audit and succession. Court readiness assembles it around authority and remedy. That is not an admission that litigation is inevitable. It is recognition that institutions with public effects must remain governable when conflict occurs.

The remedy matrix

A practical remedy matrix can align breach, target and continuity protection. For failure to report, require a sworn report, personal attendance and a short deadline while leaving services untouched. For refusal to execute a specific lawful act, authorise a substitute and preserve an audit trail. For threatened dissipation, restrain the asset or designated account while protecting essential payments. For credential obstruction, order transfer or supervised recovery rather than broad system seizure.

For repeated deliberate non-compliance, escalating personal or corporate financial consequences may be justified, but the continuity budget should remain visible. For disputed technical effects, appoint an independent expert and review quickly. For governance incapacity, a receiver or monitor may be necessary, but the mandate should specify services, powers, reporting and exit.

No matrix can decide a case automatically. Evidence of intent, ability, urgency, prior breaches and legal authority matters. The value of the matrix is to force alternatives onto the record. Before a sanction reaches shared infrastructure, the court can ask why actor-specific, transaction-specific or task-specific enforcement would not work.

The institution should present the same analysis proactively. A party that proposes a credible compliance path is more persuasive than one that only predicts catastrophe. Courts should be sceptical of continuity claims used to preserve control, but responsive to tested safeguards that protect third parties while obedience occurs.

What proof would change the conclusion

The conclusion that coercion should normally target responsible actors rather than essential services is strong but not absolute. Evidence could show that the service itself is being used to continue the prohibited act, conceal assets, alter records or frustrate the order. In that case, functional restraint may be necessary. Evidence could also show that officers are interchangeable fronts and only institutional assets create effective pressure.

Even then, the court should identify the minimum affected function and review duration. If a database is being manipulated, read-only custody may be enough. If funds support continued disobedience, designated spending control may be enough. If the whole operating chain is compromised, broader intervention may be justified, but an independent continuity operator should be considered.

Conversely, proof that a claimed service risk is remote, exaggerated or easily mitigated would weaken demands for protection. The institution bears responsibility for producing accurate dependency evidence. Past service continuity does not prove future resilience, and public importance does not prove that current management is indispensable.

This openness to contrary proof is essential. Continuity-aware enforcement is not a doctrine of deference to registries. It is a method for allocating coercion according to mechanism, responsibility and incidence.

Compliance and continuity reinforce the same legitimacy

The deepest risk in a registry crisis is not merely an outage or an adverse judgment. It is the belief that institutional power cannot be corrected without endangering the service, and therefore must not be corrected at all. That belief turns technical dependence into political immunity.

The answer is not to deny dependence. It is to design around it. Courts should require precise commands, locate control, preserve evidence, escalate against responsible actors and protect defined essential functions. Registries should maintain authority maps, substitute access, continuity budgets, vendor readiness and verified reporting. Members should receive operational facts without being recruited into litigation narratives.

AFRINIC's published record shows why this work matters. The regional registry function has continued through disputes, changes in authority and judicial intervention, but continuity cannot rest on improvisation or personalities. It needs a durable architecture that survives both disobedience and enforcement.

A court that can compel the right person without disabling the wrong network strengthens the rule of law. A registry that can obey without failing strengthens Internet coordination. Those are not competing achievements. They are the same institutional test viewed from opposite sides.