Summary

  • Cloud Innovation's 7 March motion asked for two interim orders: a restraint on five kinds of structural change at AFRINIC, and appointment of the Official Receiver to preserve the company's assets and business value.
  • The filing did not appoint a receiver. AFRINIC resisted the application, and the Commercial/Bankruptcy Division granted relief only on 12 September 2023.
  • The reported prayers addressed the company, its management, its assets, and its value. They did not ask the Court to transfer title to Internet number resources, although the full supporting affidavit is not available in the sealed record used here.
  • A court-supervised custodian can be a legitimate continuity backstop when a private technical company cannot govern itself, but the mandate must keep corporate custody distinct from registry policy, technical operations, and number-resource ownership.
  • The sound institutional response is bounded accountability: precise scope, neutral records, insulated services, auditable decisions, fixed reporting, and a legible handback to properly constituted corporate organs.

A filing aimed at the corporate shell

The starting point is the verb asked. On 7 March 2023, Cloud Innovation filed a motion paper supported by affidavit in the Bankruptcy Division of the Supreme Court of Mauritius. The later appellate judgment reports two prayers for interim relief under section 178 of the Companies Act. The first would restrain AFRINIC from relocating or subjecting itself to takeover, merger, restructuring, or management control in any manner. The second would appoint the Official Receiver to “hold the ring,” preserve the status quo of AFRINIC's assets, and maintain the value of its business.

Those requests were substantial. Together they would place a legal boundary around changes to the company and a public officeholder inside that boundary as temporary custodian. Yet they were not self-executing. A litigant's motion states the relief it wants; it does not supply the Court's decision. AFRINIC's legal advisers were resisting the application by 28 March and raised, among other matters, Cloud Innovation's standing. The Commercial/Bankruptcy Division later granted relief on 12 September. Collapsing those dates would turn an adversarial request into an accomplished judicial act five months too early.

The same discipline is necessary when describing what was to be preserved. The quoted prayers refer to corporate structure, management, assets, and business value. They do not request a transfer of title to Internet number resources. That observation is limited to the prayers as reported in the later judgment: the full motion paper and supporting affidavit are not part of the record available for this analysis. It would therefore be too broad to say the entire case said nothing about number resources. It is accurate to say that the two reported prayers did not ask for their ownership to be conveyed.

This is more than lawyerly caution. AFRINIC matters because it operates a technical registry whose records are relied upon by members and network operators. But the importance of the service does not change the legal nature of the institution. AFRINIC is a private, member-based company and coordinator. It keeps records, supports uniqueness, and performs technical administration. It is not a state and does not acquire sovereignty over the numbers it records. The Supreme Court, by contrast, exercises public judicial authority under Mauritian law over a company incorporated in Mauritius.

A court can regulate the company's affairs without adopting the company's ledger as a register of sovereign property.

The 7 March application should thus be read on the layer where it was made. It sought to prevent movement in the corporate shell and to preserve value while a court considered how the company should be held. Its seriousness lies in the possibility that corporate paralysis could compromise accountable operation of an important registry. Its limit lies in the fact that neither corporate restraint nor temporary custody answers the separate question of who owns, uses, or has legitimate operational control of Internet number resources.

Two prayers, two forms of control

The first prayer was preventive. It named five structural acts: relocation, takeover, merger, restructuring, and management control. These are not interchangeable. Relocation concerns the corporate or institutional seat. Takeover and merger concern changes in control or combination. Restructuring can alter the company's organisation, liabilities, or operating arrangement. Management control concerns who can direct the company. By listing each act, the prayer drew a perimeter around ways in which AFRINIC's legal form or decision-making centre might change while the dispute was unresolved.

The second prayer was custodial. It asked the Court to appoint the Official Receiver to hold the ring, preserve the status quo of the company's assets, and maintain business value. The language evokes conservation rather than disposal: keep the field from shifting, protect what exists, and prevent avoidable erosion. It does not describe a sale of number resources, an allocation-policy decision, a transfer of registered holdings, or a technical instruction to alter routes or RPKI. Nor does the filing itself tell us the precise operational acts a later custodian would be authorised to perform.

Read together, the prayers sought a pause and a keeper. The pause would limit structural movement. The keeper would provide a court-supervised point of custody while ordinary corporate governance was in dispute. This pairing is intelligible in company law. If a company cannot reliably speak or act through its normal organs, merely prohibiting change may preserve paralysis. Merely appointing a custodian without a structural restraint may leave the perimeter unstable. The requested combination attempted to maintain both the entity and its value while the Court addressed the conflict.

But “hold the ring” is not a complete operating constitution. It does not, by itself, define the line between preserving corporate records and changing the substantive interests recorded in them. It does not identify which technical functions must continue untouched, which decisions require notice, how conflicts are handled, or when custody ends. Those questions become especially important when the company operates infrastructure-adjacent services. The two prayers describe the architecture of interim corporate protection; they do not resolve every consequence of putting that architecture around a live registry.

The precise control surface therefore begins with what the applicant was entitled to allege, continues through the Court's just-and-equitable discretion, and ends with the scope of any order actually made. It includes the named structural restraints, the identity and legal position of the proposed receiver, the categories of assets and records to be preserved, the duration of custody, reporting duties, protection of technical services, and the route back to functioning corporate organs. On 7 March, most of that surface remained a proposal awaiting judicial determination.

Why request and order cannot be merged

Public accounts of institutional crises often compress procedure into outcome. An application is described as if the requested remedy already existed; a later order is read backward into the filing; the existence of a court case becomes evidence that one side's characterisation was accepted. In this matter, each compression distorts the central event.

Cloud Innovation was the applicant. That gave it the ability to put allegations and requested relief before the Court, subject to the legal requirements governing access to section 178. It did not make Cloud Innovation the custodian, the owner of AFRINIC, or the controller of registered Internet number resources. AFRINIC's resistance confirms that the application was contested, not a joint proposal. The Court remained the institution with authority to decide whether relief was justified and on what terms.

The timing also protects the integrity of causation. No effect can properly be attributed to a receivership on 7 March because no receiver was appointed by the act of filing. The record used here proves no outage, route change, RPKI change, registration change, or quantified loss resulting from the motion. Even after acknowledging that relief was granted in September, it would be wrong to use the later fact of appointment as proof that the March filing itself changed technical operations.

Keeping the dates apart does not minimise the application. A credible request for structural restraint can influence planning, legal risk, counterparties, and confidence before a decision. Members and operators may need to ask who can sign, who can instruct staff, or whether a contemplated corporate step could be restrained. Those are real diligence costs. Yet they are different from a judicial transfer of authority and different again from an observed change in the technical registry.

The clean formulation is simple. On 7 March, Cloud Innovation asked. AFRINIC resisted. On 12 September, the Commercial/Bankruptcy Division later granted the corporate restraint and appointed the Official Receiver. This article does not turn the later receiver's tenure into the subject of the March motion, nor does it treat the later order as evidence of number-resource ownership. The distinction preserves both the applicant's actual act and the Court's separate authority.

What section 178 offered—and what it did not decide

Section 178 supplies a company-law route for specified applicants to seek judicial relief from oppressive, unfairly discriminatory, or unfairly prejudicial conduct or acts. The statutory doorway matters because a court does not obtain authority from the technical importance of AFRINIC or from the preferences of the registry community. Its authority comes from Mauritian law governing a Mauritian company. That is the public-law anchor for the requested intervention.

The provision gives the Court a just-and-equitable remedial discretion. Its listed tools include regulation of the future conduct of a company's affairs at paragraph 178(2)(c) and appointment of a receiver at paragraph 178(2)(e). Those categories make the structure of Cloud Innovation's two prayers legible. One sought future restraint on specified corporate conduct; the other sought a receiver to preserve the company while the restraint operated.

The statute does not establish the allegations merely because an application cites it. It does not itself prove that Cloud Innovation fell within an eligible applicant class, settle the standing objection, decide that relief was just and equitable, or appoint anyone. Nor does the statute decide ownership of Internet number resources. Its subject here is the affairs and remedies of a company. The fact that a company's records concern valuable operational resources does not silently expand every company-law remedy into a property judgment about those resources.

One detail in the surviving instruments deserves to remain visible. The scanned 12 September order labels both the restraint and the receiver appointment under section 178(2)(C). The statutory text, however, lists regulation of future conduct at paragraph (c) and appointment of a receiver at paragraph (e). The later appellate judgment summarises both March prayers as made under section 178 but does not resolve this textual mismatch. The discrepancy should neither be edited away nor promoted into a conclusion the sources do not provide. It is a citation issue in the record.

Without further authoritative treatment, it cannot responsibly be said to validate or invalidate the relief.

That restraint illustrates a broader point about institutional analysis. Precision is not pedantry when several forms of authority are adjacent. Applicant standing, judicial discretion, statutory remedy, corporate custody, and technical administration are separate links. If they are fused, a reader may conclude that a company-law application gave a private party power over public communications infrastructure.

If they are kept separate, the event is easier to understand: an applicant invoked an ordinary legal mechanism; a court later exercised state authority; the regulated object was a private company; and the technical significance of that company created continuity concerns without changing the source of legal power.

Corporate assets are not automatically the things in the ledger

The phrase “assets of AFRINIC” can carry more weight than it can safely bear. Corporate assets may include money, equipment, contractual rights, intellectual property, records, claims, and other interests recognised by law. Business value can include the company's ability to continue operating, the organisation of its staff and systems, relationships with members and counterparties, and the integrity of records needed to deliver its service. The prayers used broad preservation language, but broad corporate language should not be converted into an unproved catalogue of ownership.

Internet number resources introduce a distinctive semantic trap. AFRINIC maintains registration and coordination records associated with IPv4, IPv6, and autonomous system resources in its service region. A registry's database can record who is recognised for administrative or operational purposes. The organisation can have contractual and policy relationships surrounding those records. None of that means the registry owns every number recorded, any more than maintaining a ledger makes the bookkeeper the beneficial owner of every interest entered in it.

This difference remains important even if registry information is operationally consequential. Accurate records help networks coordinate. Errors, delays, or disputes can increase risk. The registry's service therefore has real power in the ordinary sense of affecting what participants can do. But practical influence is not sovereignty, and record administration is not title. Calling the database a corporate asset may describe the company's possession of systems, software, or compiled records; it does not answer the legal nature of each recorded number-resource interest.

The two March prayers, as reported, did not ask the Court to decide that question. They sought to restrain changes to AFRINIC and preserve the status quo of its assets and business value. An order answering those prayers could place corporate materials in custody without transferring beneficial title in everything referenced by those materials. It could preserve a database while leaving the substantive relationships recorded within it unchanged. It could keep the company capable of serving members without making the custodian an owner of members' operational interests.

This is the correct burden of proof. Anyone claiming that the motion transferred Internet-number ownership would need a source that actually establishes a conveyance or adjudication. The available record does not. Anyone claiming that the filing changed a registry record, a route, or RPKI would need evidence of that change and its causal link. None is proved here. The absence of such proof is not a declaration that no later operational event ever occurred; it is a boundary on what can be said about the 7 March motion.

For a technically dependent audience, the safest inventory is therefore layered. Corporate assets should be identified as corporate assets. Legal records should be identified as legal records. Technical registry data should be identified as data necessary to perform a coordination service. Number-resource control records should be distinguished from any asserted beneficial or contractual interest. Live operator dependencies should be mapped without implying that the company owns the networks that rely on accurate administration. A receiver may be asked to preserve one layer without absorbing all the others.

A private registry and a public court

AFRINIC's role is easier to defend when it is described without mythology. It is a private member-based technical registry, a coordinator, and a service provider. Its legitimate work includes maintaining accurate records and helping preserve uniqueness and interoperability. Those functions require expertise, consistency, and trusted procedures. They do not make AFRINIC a legislature, regulator, police force, prosecutor, court, confiscator, or continental sovereign.

That limit does not make AFRINIC unimportant or powerless. Private institutions can control access to services, maintain decisive records, enforce contracts, and make policy choices that impose meaningful costs. Members may depend on them. Operators may build processes around their reliability. The right response to that practical influence is accountable governance and legally intelligible authority, not the invention of sovereignty.

The Supreme Court occupies a different category. It exercises the state's judicial power. When it acts under the Companies Act, it can bind a domestic legal person because Mauritian law gives it that authority. The Court does not need AFRINIC to be a public authority in order to regulate AFRINIC's corporate affairs. Conversely, the Court's intervention does not elevate the registry's private coordination into public government.

Public legal authority remains with the Court; technical administration remains a function of the company or a properly authorised custodian; the recorded resources remain outside any ownership inference not established by law and evidence.

This arrangement is ordinary in one sense and unusual in another. It is ordinary because companies that perform important services remain subject to the law of their incorporation. A private institution cannot place itself beyond company law by describing its mission as regional or community-based. It is unusual because defects in the governance of this company can create concern far beyond its office. The corporate dispute sits next to a technical system on which many actors rely, so loose words about control can be mistaken for changes in the network itself.

The temptation is to solve that discomfort by calling AFRINIC a steward of a public resource and then treating its internal processes as a kind of constitution. But that language obscures more than it clarifies. Community participation may be useful. Technical policy may be widely accepted. Membership rules may be binding by contract. None produces sovereign authority. Ordinary law, jurisdiction, liability, and enforceable duties remain the mechanisms for addressing a private company's inability to govern itself.

The 7 March motion is significant precisely because it brought that reality into view. Cloud Innovation did not need to ask a registry community to create a sovereign remedy. It asked a state court to use a statutory corporate remedy. The Court's capacity to consider the application arose from AFRINIC's status as a Mauritian company. The technical importance of AFRINIC made the design of relief more consequential, but not more metaphysical.

The strongest case for a temporary custodian

The benign case for the requested relief deserves to be stated at full strength. A private technical company can reach a point at which its normal organs cannot reliably make decisions, authorise action, or protect value. At that point, insisting that ordinary governance continue unaided may be formalism rather than restraint. If no responsible actor can lawfully act for the company, staff, records, contracts, and services can be exposed to drift. A court-supervised receiver can provide a recognised point of accountability while lawful governance is restored.

The structural restraint also has a coherent protective logic. Preventing relocation, takeover, merger, restructuring, or management control can stop the legal perimeter from moving while the Court examines the dispute. Without such a restraint, the identity or control of the company might change before the application can be resolved. The relief can therefore preserve the Court's ability to give an effective judgment and prevent corporate value from being dissipated through contested acts.

For a registry, continuity strengthens this case. Operators do not benefit from uncertainty about who can maintain essential systems, pay staff, preserve records, or honour contracts. Members need an entity capable of receiving instructions and reporting on its affairs. Counterparties need to know which signature is valid. A temporary custodian, accountable to a court rather than to one litigant, can reduce the risk that competing claims leave the company unable to perform routine duties.

Nothing in that account is an attack on the Internet. Nor is court supervision necessarily hostile to the registry model. It can be the mechanism by which a private technical institution is kept within its actual legal form and returned to functioning organs. The public interest in stable coordination may be served by preventing abrupt corporate change, preserving the evidence needed to reconstruct decisions, and sustaining ordinary service while the conflict is adjudicated.

The phrase “maintain the value of the business” also has a legitimate continuity dimension. A functioning organisation is more than a bank balance. Its value can depend on staff knowledge, orderly records, service relationships, procedural credibility, and the ability to operate without contradictory commands. Preserving those elements can protect members and operators as well as creditors or litigants. A receiver's accountability may be preferable to an institutional vacuum.

This benign account does not prove that every requested term was necessary or that every later act was successful. Those questions lie beyond the evidence and scope here. It establishes why a court could rationally be asked to consider temporary corporate custody. The answer to overreach cannot be a refusal to recognise paralysis. It must be a design that captures the benefits of custody while limiting what custody means.

The answer: make the mandate exact

The best answer to the benign case begins by accepting its premise. Sometimes a company needs an external, lawful caretaker. The danger arises when a mandate to preserve the company is expressed or narrated as authority over everything the company touches. In a registry setting, that rhetorical expansion can move quickly from corporate files to registry data, from registry data to number resources, and from number resources to claims of technical sovereignty. Each step requires authority and evidence of its own.

An exact mandate would identify the legal objects in custody. It would distinguish company property from member information, contractual records from technical control records, software and infrastructure from the operational interests represented in the database. It would say what preservation allows: routine maintenance, security work, payment of essential obligations, staff supervision, and protection against unauthorised alteration may be necessary.

It would also say what preservation does not allow without further approval: changing substantive registration interests, redesigning policy, disposing of contested assets, or using temporary custody to settle long-term governance choices.

The mandate would also separate roles. The Court supplies public authority. The receiver performs delegated custodial tasks. AFRINIC remains the private technical institution whose service must be maintained or restored. Members remain participants with contractual and governance interests, not a substitute state. Operators remain dependent users of reliable coordination, not subjects of a new jurisdiction. Cloud Innovation remains the applicant, not the beneficiary of every power the Court might entrust to a neutral officeholder.

Time is as important as scope. Temporary custody should have reporting dates, objective completion conditions, and a clear path to handback. An open-ended caretaker arrangement can become an alternative governance system simply because no one knows when it ends. A condition-based exit—tied, for example, to properly constituted corporate organs capable of receiving records and resuming lawful management—keeps preservation directed toward restoration.

Technical neutrality needs its own protection. Routine services should continue under documented procedures, with heightened review for any action that changes registration, authentication, routing-security support, or related operational state. An auditable log should distinguish maintenance from discretionary policy and emergency intervention from ordinary administration. Privacy-safe reporting can show members and operators what class of action occurred without exposing sensitive data.

Finally, the wording should protect against false ownership implications. Preserving a record is not acquiring the interest recorded. Maintaining a registry is not becoming owner of its entries. Court supervision of a company does not make the Court an Internet allocator. The clearer those propositions are at the start, the less likely a necessary corporate remedy is to produce costly ambiguity for the people who depend on the service.

How corporate ambiguity becomes operational cost

The sealed record proves no outage or technical change from the March filing. The credible impact mechanism is more indirect. When a company's authority structure is uncertain, everyone dealing with it must spend more effort establishing who can act, which decisions are valid, and whether records will remain reliable. In an ordinary business, that may delay contracts and payments. In a technical registry, it can also affect confidence in administrative processes that operators expect to be stable.

Consider the decisions a network operator may need to make without assuming any actual service failure. Legal and engineering teams may review who is authorised to receive requests, sign agreements, approve changes, or attest to the integrity of registry processes. They may preserve additional evidence, delay non-essential changes, seek alternative contacts, or ask whether an action taken during the dispute will later be challenged. Each step consumes time and professional attention. The cost is a burden of verification, not proof that packets stopped moving.

Members face a related problem. A member-based institution ordinarily derives practical legitimacy from rules, elections, boards, officers, and accountable administration. If those organs cannot operate, members need to know which parts of participation remain meaningful and which powers a temporary custodian is performing. Silence encourages two opposite errors: some observers may assume the receiver has unlimited power; others may assume no valid action is possible. Transparent, bounded reporting narrows that range of speculation.

Staff and counterparties also need a reliable chain of authority. Employees must know whose instructions are lawful. Banks, landlords, suppliers, and service providers need valid approvals. Data custodians need a basis for granting or restricting access. None of these questions requires the receiver to own Internet numbers. They require the company to remain capable of acting through a legally recognised channel.

The most damaging ambiguity is a category mistake presented as reassurance. If officials say that control of the company means control of the region's numbers, they may intend to signal continuity. Instead, they inflate the institution's role and invite every corporate contest to be read as a contest for Internet sovereignty. If they say that the corporate dispute has no relevance because the Internet is decentralised, they ignore the real dependencies built around registry records and services.

The accurate position is between those extremes: the registry is not sovereign, but failures in its governance can impose operational and legal costs.

Corporate-institution stability and network continuity must therefore be measured separately. The first concerns the legal entity, its organs, assets, records, staff, and business value. The second concerns the continued ability of networks and customers to rely on routing, addressing, and associated services. Stabilising the company may support continuity, but the two outcomes are not identical. A company could be legally preserved while service quality deteriorates; technical services could continue for a time while corporate authority remains contested. Policy should monitor both rather than treating one as a proxy for the other.

This distinction disciplines causal claims. The March prayers show an attempt to stabilise corporate affairs. They do not prove that a route changed or did not change, that RPKI was affected or unaffected, or that a customer suffered loss. Those propositions require their own operational evidence. In its absence, the responsible claim is that ambiguity creates exposure: it raises diligence costs and makes it more important to define who can touch which systems under what authority.

The evidence architecture and its limits

The central facts come from official legal materials, but each material answers a different question. The later Court of Civil Appeal judgment reports that the motion was filed on 7 March, identifies its affidavit support, and quotes the two prayers. It also records later resistance and the eventual procedural history. Because it is a later judgment, it is evidence of how the March application was reported by the Court, not a substitute for the complete motion paper or affidavit.

The scanned September order proves that the Commercial/Bankruptcy Division later granted the restraint and appointed the Official Receiver. It does not convert 12 September into 7 March, and its redactions should not be filled by inference. It is useful for confirming the disposition and the preservation language, not for reconstructing missing parties, reasons, or operational details.

The Companies Act establishes the statutory structure. It tells us who may apply under section 178, the kind of conduct or acts that open the remedial doorway, the Court's just-and-equitable discretion, and the menu of possible orders. It does not prove Cloud Innovation's allegations, resolve standing, or decide what any particular corporate record represents. A statute can authorise appointment of a receiver without making every subject touched by the company an asset owned by that receiver.

Institutional analysis then supplies the distinctions needed to avoid overreading the legal instruments. AFRINIC's private member-based form explains why company law can reach it and why member accountability matters. The bookkeeper distinction explains why maintaining records is not sovereign authorship of the interests recorded. Operator-continuity analysis explains why corporate uncertainty can matter without assuming a realised outage. Prior accounts of the judicial history help locate this episode, but they do not replace the court record for the exact prayers.

This hierarchy matters because analytical sources should not be made to prove procedural facts, and court papers should not be made to prove a philosophy of technical authority they do not articulate. The filing date and wording of the prayers belong to the judicial record. The statutory categories belong to the Act. The claim that record-keeping is not ownership is an institutional and legal distinction applied to those facts. The lack of evidence for route or registry changes remains a lack of evidence; it is not filled by a general essay about registry risk.

The narrowness of the evidence also places limits on negative claims. The reported prayers do not seek title to number resources. That supports a precise absence statement about their text. Because the complete affidavit is unavailable here, it would be improper to claim that no part of Cloud Innovation's case discussed number resources. Likewise, the sources do not prove a transfer. That does not establish a universal proposition about every legal interest associated with every registry entry. It means this motion cannot be cited as proof of a transfer that its reported relief did not request.

The result is an evidence ledger with deliberate blank spaces. Motive is blank. Bad faith is blank. Any alleged corruption, capture, or concealment is blank. Quantified economic harm is blank. Technical changes caused by the filing are blank. The later receiver's performance is outside the frame. A rigorous article does not fill those spaces with suspicion simply because the underlying dispute was contentious.