Summary

  • In SC/COM/WRT/000056/2023, Larus Cloud Service Limited asked the Supreme Court of Mauritius for urgent relief against African Network Information Centre (AfriNIC) Ltd and Executive Services Limited; the judge intervened before the respondents had been served.
  • The interim order restrained effect or further effect to a written directors’ resolution dated 17 January 2023 and restrained further written resolutions in the absence of a quorate board. The listed subjects were the 2023 budget, two separately worded interim-executive appointments, the internal auditor, Central Africa Seat 4 and an elected Governance Committee member—control points for spending, administration, assurance, board representation and oversight.
  • The case must not be folded into Cloud Innovation Ltd’s resource-registration litigation. This instrument names a different applicant and addresses corporate governance acts, not an RSA, address blocks, leasing, deregistration, RPKI, WHOIS or resource recovery.
  • Because the order was ex parte and interim, it proves neither that the resolution was finally unlawful nor that the board was finally found to lack quorum. Service, a 14 February 2023 return date, a show-cause process and Larus Cloud’s undertaking in damages made the restraint temporary and reviewable.
  • AFRINIC remained a private member-based technical registry, bookkeeper, service provider and coordinator. The Court and ordinary law supplied sovereign authority; the successful applicant gained relief, not power to govern the registry.

A judicial pause with six pressure points

The first useful way to read the order is not as a victory banner but as a pause button. The document is only two pages long. It records that Larus Cloud Service Limited came before the Commercial Division of the Supreme Court of Mauritius, in Chambers, with an application, a praecipe, an affidavit dated 31 January 2023 and annexed documents. Judge P. D. R. Goordyal-Chittoo considered the matter sufficiently urgent to justify immediate intervention before notice of the application had been served on the respondents. That posture matters. One side had been heard; the other had not yet had its opportunity.

The pause nevertheless reached important ground. The Court restrained African Network Information Centre (AfriNIC) Ltd and Executive Services Limited, together with their agents, preposes or representatives, from acting upon or giving effect or further effect to a document described as the Written Resolution of Directors Passed on 17th January 2023. It also restrained them from passing written resolutions in the absence of a quorate board of directors. That is the order’s expressed legal action. It did not appoint anyone, dismiss anyone, approve a rival budget or install Larus Cloud as a corporate decision-maker.

Attached to the named resolution, as described by the Court, were six agenda items. They were approval of the 2023 budget; appointment of an Officer-in-Charge as Interim Chief Executive Officer; appointment of an Internal Auditor; appointment of an Interim Chief Executive Officer; appointment of Central Africa — Seat 4; and appointment of an Elected Governance Committee Member. The overlap in wording between the two executive items is conspicuous, but the published instrument does not explain it. Any confident story about duplication, succession, intended personnel or motive would outrun the record.

Within the first few lines of analysis, then, three boundaries have to remain visible. The order was urgent but provisional. It restrained acts rather than reallocating their underlying powers. And the six subjects were corporate control points, not technical registry entries. Those boundaries are not lawyerly decoration. They determine what the event can teach us about authority.

A budget is more than a spreadsheet. It can authorise payroll, vendor commitments, service capacity and ordinary obligations. An executive appointment can determine who instructs staff, signs documents or directs spending. An internal auditor can influence the quality of internal control and the credibility of reporting. A board seat and a Governance Committee position can shape votes and oversight. None of that proves that a particular act was wrongful, that funds were mishandled or that any service failed. It does show why uncertainty about who may decide can quickly become uncertainty about what the organisation may do.

That is the central institutional mechanism in this case. A member-facing private company invoked a sovereign court’s ordinary process to put a temporary hold on specified acts of another private company. The relief came with a route to service and adversarial review. It did not arise from a supposed power of the “Internet community,” from regional identity or from control of the registry database. It arose because a court could examine the corporate authority claimed for a resolution and preserve the position pending further process.

The name in the caption changes the meaning

AFRINIC’s disputes are often flattened into one long confrontation. That habit is particularly misleading here. The applicant named in the order is Larus Cloud Service Limited. It is not Cloud Innovation Ltd. The respondents are African Network Information Centre (AfriNIC) Ltd and Executive Services Limited. The relief described in the order concerns a written directors’ resolution and the absence of a quorate board. The document contains no claim about IP address blocks, a registration-services agreement, leasing, out-of-region use, deregistration, resource recovery, RPKI or WHOIS.

AFRINIC has publicly characterised Larus Cloud and other companies as associated with Cloud Innovation. An institutional self-description or a party’s litigation narrative can be relevant to understanding how that party catalogued events. It cannot erase separate legal identities. Association is not merger. A caption is not a family nickname. Each company bears its own rights, obligations and potential liabilities; each application must be read against the relief it actually requests and the instrument the Court actually issues.

This distinction does more than avoid a technical error. It produces the article’s central insight. If the January 2023 order were merely another step in a resource-registration case, it would tell us little beyond the persistence of that controversy. Read on its own terms, it shows a different control surface. Judicial supervision had reached a member-facing dispute over how AFRINIC made corporate decisions: how a budget might be approved, how executive and audit roles might be filled, how a regional seat and committee role might be handled, and whether written resolutions could proceed without a quorate board.

The difference is institutional, not rhetorical. Resource litigation asks about contractual and registry treatment of number resources. This order’s expressed target was corporate authority. It therefore demonstrates that court scrutiny of AFRINIC was not confined to one company’s resource dispute. A separate applicant could ask ordinary law to review the basis on which governance acts were being carried forward. That does not tell us whether the applicant ultimately prevailed. It tells us that the registry’s corporate machinery was answerable to a forum outside the registry’s own claims about mandate or community.

Nor should the opposite error creep in. Distinguishing the case from Cloud Innovation does not make it wholly unrelated to a wider climate of institutional strain, and it does not prove that every contention advanced by Larus Cloud was sound. The responsible claim is narrower: the legal identity, relief and documentary target are distinct, and those distinctions matter. Anything more would require the application, affidavit, annexures, respondents’ evidence and later decisions, none of which appears in the published record used here.

What an interim order can—and cannot—carry

Urgent ex parte relief is powerful precisely because it can operate before the other side has been heard. That is also why a careful reader should resist turning it into a merits judgment. The Court recorded urgency and imposed an interim restraint. The instrument does not finally declare the 17 January resolution void, invalid, unlawful or fraudulent. It does not finally determine the board’s composition or hold, after adversarial evidence, that the board lacked quorum. It does not resolve the intended main action or award damages. It does not disclose what happened on or after the return date.

The strongest caution is simple: at the ex parte stage, the judge had material from the applicant, while the respondents had not yet been served. We do not have the application, affidavit or annexures, and we do not have the response. We cannot test the competing account, inspect the director register, calculate quorum from the operative facts or know what legal arguments would later be advanced. A temporary judicial choice to preserve a position is not a final endorsement of every allegation used to request it.

The order itself protects against that inflation. It directed that the respondents be served. It invited them to show cause why the interim order should not be made interlocutory pending determination of a main action Larus Cloud intended to file. Their position was to be communicated through the e-filing system on or by Tuesday 14 February 2023, when the matter was returnable. That sequence converts an emergency intervention into a reviewable process: first a short-term restraint, then notice, response and reconsideration.

The duration language deserves equal care. The order said that the interim injunction would remain in force until determination of the present matter. It also referred to the possibility of interlocutory relief pending a main action. Those phrases show a procedural bridge, not a final destination. The Court was maintaining control over the next step. Without the later record, it would be wrong to supply an outcome, infer that an interlocutory order followed or treat the restraint as permanent.

The applicant’s undertaking gives the bridge a load-bearing support. The injunction issued at Larus Cloud’s own risks and perils and on an undertaking through counsel to abide by any later order concerning damages caused to the respondents by the interim relief. That language recognises a basic asymmetry. An applicant may be permitted to interrupt corporate acts before the respondents are heard, but the interruption can itself cause harm. The possibility of later liability helps couple urgent power with consequence.

Calling that undertaking “boilerplate” would miss its institutional function. It does not prove that compensable loss occurred, set a quantum or guarantee recovery. It does mean the applicant was not granted a costless veto. Temporary relief carried exposure to the same ordinary legal system that supplied the remedy. In authority terms, Larus Cloud was a litigant asking a court to act, not a superior directing AFRINIC. The Court retained control of the order, and the undertaking left room for a later judicial response if the restraint caused damages.

The strongest benign case for action

There is a serious argument on the other side of any injunction touching a technically important organisation. AFRINIC needed to function. An organisation confronting unsettled leadership could still require a budget, executive administration, internal audit, board representation and governance oversight. Vendors might need instructions; employees might need a clear chain of responsibility; technical and contractual services might need continuity. Written resolutions can be an efficient corporate tool, especially when directors cannot easily convene in one room.

Freezing such acts can therefore create its own risks. A delayed budget may constrain routine commitments. Uncertainty over executive authority can slow decisions. An unfilled audit role can weaken assurance. Gaps in a board or committee can make future oversight harder. For a registry on which operators depend, prolonged institutional paralysis is not an abstract inconvenience. The narrow technical ledger may require stable staffing, security attention, support and coordinated updates even while the company’s leadership is disputed.

This is the strongest benign account because it treats continuity as a real public-facing concern without pretending that AFRINIC is a public sovereign. Technical dependence can increase the organisation’s duty to act neutrally and keep essential services available. It does not grant the company immunity from corporate law. Nor does operational urgency make any particular written resolution valid. The need to pay a vendor cannot answer whether the directors authorising payment had power to act; the need for leadership cannot by itself establish who may appoint a leader.

The answer to the benign case is therefore not casual judicial immobilisation. It is precision. A court should understand which corporate acts are contested, which essential technical functions cannot safely pause, what authority supports each necessary act and which measures preserve the status quo without deciding the merits. The January order was targeted at a named resolution and further written resolutions in the stated quorum circumstance. It did not purport to take over AFRINIC’s database, perform registry services or manage operations.

That same precision disciplines critics of the order. One cannot infer from the injunction that a particular outage was averted or caused. The document supplies no budget amount, customer impact, employee consequence, service interruption, resource freeze, RPKI change, WHOIS change or quantified loss. Operator-continuity analysis explains why governance can matter downstream; it does not turn a plausible mechanism into an observed event.

A registry ledger is not a throne

AFRINIC’s technical role is important and narrow. It is a private member-based registry, bookkeeper, service provider and coordinator. It maintains records and services that help preserve the uniqueness and usability of Internet number resources. Networks rely on coherent registration facts, contractual administration and technical coordination. Those functions can become a critical dependency for operators across Africa.

Critical dependency, however, is not sovereignty. AFRINIC has zero sovereign, legislative, regulatory, police, prosecutorial, punitive, confiscatory or public-law adjudicatory authority. It cannot create those powers by invoking a region, a community or its own database. Its service region is not a polity. Its participants are not a demos. Its board is not a legislature. Membership is not political allegiance. Corporate documents may bind within their lawful private scope, but they do not become public law merely because the organisation performs a specialised function.

The bookkeeper analogy clarifies the boundary. A reliable bookkeeper coordinates authoritative-looking records about transactions or entitlements, but keeping the ledger does not make the bookkeeper the owner of everything recorded. In the same way, an RIR’s record helps networks coordinate. Control of the recording process can exert great practical influence, which is why neutrality, due process and continuity matter. Yet practical influence does not turn record maintenance into an inherent right to punish, confiscate, legislate, police or adjudicate.

This distinction also protects AFRINIC. If the registry were treated as a sovereign, every technical action would be burdened with political claims that a private company is not institutionally equipped to resolve. Competing groups could present themselves as “the community” and demand obedience as if participation created constitutional authority. A narrower conception—private coordination under contract and company law—provides a more defensible foundation. It asks who is authorised, under what instrument, subject to what review and bearing what liability.

In the January 2023 event, the sovereign actor was plain: the Supreme Court of Mauritius acting through ordinary law. The Court could issue a binding order to the corporate respondents within its jurisdiction. Larus Cloud could apply for relief and accept potential liability in damages. AFRINIC and Executive Services Limited could be restrained, served and heard. Those are distinct legal positions. None made Larus Cloud the principal of AFRINIC; none made the Court an operator of the registry; none enlarged AFRINIC into a regional government.

This authority map should be kept even when one approves of the result. A litigant who obtains an injunction has persuaded a court to act provisionally. It has not inherited the court’s sovereignty. It cannot use the order to approve AFRINIC’s budget, appoint officers or claim a mandate over members. The restraint stopped effect being given to specified acts; it did not transfer those acts into the applicant’s hands.

Principal, agent and the discipline of loss

Corporate authority must run from a competent source to the person exercising it. A director’s title is not a free-standing licence; a board’s power depends on the company’s constitution, applicable law, valid composition and the procedures required for the act. An executive acts within delegated authority. A committee performs the role assigned to it. Members may have rights, but membership alone does not turn any member into the corporation’s principal for every purpose.

That is why quorum is more than a counting puzzle. It is one device for deciding when a collective body may speak as the body rather than as an assortment of individuals. The published order does not give us the director register or the facts needed to reach a final conclusion on quorum. Its wording nonetheless identifies the authority problem: further written resolutions were restrained in the absence of a quorate board. The Court did not need to pretend that the registry’s technical importance answered that corporate question.

Authority is more credible when coupled to the actor who bears the consequences of misuse. The undertaking in damages embodies that principle for interim relief. Larus Cloud obtained a temporary interruption but remained exposed to a later order if the interruption caused compensable harm. The Court exercised sovereign authority and retained responsibility for supervising the remedy. The respondents were bound in the short term but entitled to be served and to contest continuation. Each role carried a different connection between decision and loss.

The same discipline should apply inside the registry. A budget proposer should identify the authority to propose and the decision-maker authorised to approve. An officer making appointments should record the delegation and effective date. Directors should document participation, conflicts, recusals and votes. Those responsible for technical continuity should be empowered to preserve narrowly defined essential services without using an emergency as a route to settle disputed governance questions.

None of this tells us that the 17 January resolution lacked those records. The resolution itself is absent. So are the annexures, board register and respondents’ account. The point is architectural: when authority and liability are documented together, a court can review a dispute without reconstructing the institution from claims made in an emergency application. Members can understand where a challenge belongs. Operators can distinguish a corporate contest from a threat to the technical ledger.

What official and institutional sources can prove

The Supreme Court order is the controlling instrument for this event. It fixes the caption, case number, date, urgent posture, restraints, six agenda items, duration language, return date and undertaking. Its silence also matters. If the two-page instrument does not reproduce the written resolution or decide the merits, a commentary source cannot quietly fill those gaps and attribute the result to the Court.

AFRINIC’s official case list has a smaller role. It can show how AFRINIC catalogued the matter at the captured time, including its description that an injunction application was granted. It cannot convert a party’s label into a judicial holding or supply a later procedural history. AFRINIC’s bylaws can establish written corporate rules and role definitions. They cannot prove actual board composition, compliance, motive or what evidence the judge accepted in this particular application.

First-class institutional commentary serves bounded purposes. Research on AFRINIC’s quorum and bylaws can illuminate why corporate composition and written rules matter. NRS can explain the member-registry relationship and operator dependence. LARUS can identify ways governance uncertainty may create infrastructure risk. Heng Lu’s analysis supplies the authority framework: a registry is a bookkeeper rather than a sovereign, and agents must trace their power to a competent principal. None of these sources decides SC/COM/WRT/000056/2023.

That hierarchy prevents two common distortions. The first is deference to institutional self-description: because AFRINIC says it represents a region or community, the argument goes, its statements carry public authority. They do not. Official material is excellent evidence of what AFRINIC recorded, said or did; it is not evidence that the organisation possesses powers ordinary law never granted. The second distortion is the use of commentary to make facts more dramatic than the instrument allows. Context can explain stakes, but only the order can prove the order.

The six acts as a map of member accountability

Taken together, the six agenda items trace the operating skeleton of a company. The budget concerns resources. The officer-in-charge and interim chief executive items concern executive command. The internal auditor concerns assurance. Central Africa Seat 4 concerns board composition. The elected Governance Committee member concerns oversight. The order does not say how those items related to one another, whether every appointment had been implemented or who was proposed. But their categories reveal why the dispute was member-facing.

Members do not become sovereign by joining a registry. They do, however, have legitimate interests defined by company law, the constitution and their membership relationship. Those interests include knowing which body may act, how its authority is demonstrated and where a contested decision can be reviewed. When core appointments and a budget are advanced by written resolution, the validity of the decision route is not merely an internal administrative detail. It affects whether the company can credibly bind itself.

The judicial route is important because it avoids two false choices. Members need not seize the registry or proclaim a rival “community” in order to challenge authority. And the registry need not treat criticism as an attack on its technical mission. A court can ask a narrow corporate question, issue temporary relief when justified, hear the other side and later revise or discharge that relief. The technical coordination role can continue within a protected scope while the legal dispute remains legal.

The order’s member-facing quality also explains why it expands the map of judicial supervision. A resource-registration dispute can be portrayed, rightly or wrongly, as an argument at the edge of AFRINIC’s service contract. A challenge to written governance acts reaches the machinery by which the organisation itself decides. The Court’s intervention did not decide who should run AFRINIC. It showed that the claimed method of running it could be brought before ordinary law.

A better counterfactual: separate the corporate file from the technical ledger

Imagine that every urgent challenge arrived with two clean, current packets. The first would be a corporate-authority record. It would identify every director, vacancy and term; reproduce the operative constitution; calculate quorum for the relevant date; record conflicts and recusals; state the legal basis for a written resolution; show circulation, votes and implementation status; and identify the authority for each budget, appointment, seat or committee action. It would also catalogue any overlapping court orders and who was responsible for compliance.

The second would be a technical-continuity schedule. It would define the minimum work necessary to preserve ledger uniqueness, WHOIS, RDAP, reverse DNS, RPKI, contractual support, security operations and member assistance. Each function would have a named, limited authority that did not depend on winning the wider governance contest. Spending needed for those functions would be identified and controlled. Emergency actions would be logged and reviewable. The schedule would not decide who ultimately governs; it would keep neutral coordination from becoming leverage.

With those packets, a judge could examine the authority for the disputed acts without fearing that every question about a director’s mandate threatened the registry’s technical survival. The parties could distinguish acts that preserve the status quo from acts that reshape control. Members could see a lawful path for challenge. Operators could evaluate continuity risk from documented responsibilities rather than rumours about who controlled the organisation.

This design is better than granting AFRINIC a larger mandate. Enlarged claims of regional or community authority would make accountability harder because there would be no competent sovereign principal behind them. It is also better than letting a successful litigant supervise the registry. A member may invoke the law, but it should not acquire the very powers it challenges. The durable arrangement is narrower: corporate authority documented under ordinary law, judicial remedies constrained by procedure and liability, and technical continuity insulated from the contest.

The meaning of the pause

The 31 January 2023 injunction did not crown Larus Cloud. It did not regulate the Internet. It did not establish wrongdoing, finally invalidate a resolution or pronounce a final verdict on quorum. It was an urgent, temporary restraint issued before the respondents had been heard, with service, a return date and an undertaking in damages built into the path forward.

Its importance is nonetheless substantial. A separate member-facing applicant had placed specified AFRINIC governance acts before the Supreme Court of Mauritius. The targets were a budget, executive and audit appointments, a regional board seat, a Governance Committee member and the use of written resolutions under a stated quorum condition. That record shows judicial supervision reaching beyond the Cloud Innovation resource controversy into the registry’s corporate decision machinery.

The correct lesson is not that courts should manage registries or that registries should acquire public power to resist courts. It is that a private technical dependency needs unusually disciplined boundaries. The Court supplies sovereign adjudicatory authority. The company supplies corporate records and obeys ordinary law. Members use lawful avenues of challenge. Operators receive neutral technical continuity. The registry keeps the book; it does not ascend a throne.