The member who cannot shop elsewhere
A small data-centre operator in Accra, Nairobi or Port Louis does not experience AFRINIC as a debating society. It experiences AFRINIC as the place where its network identity becomes legible. The operator pays fees, keeps contact data current, asks for reverse DNS changes, maintains public registry objects, depends on routing-security publication, and needs a recognised record when a bank, customer, upstream, auditor or counterparty asks who controls a block of addresses. If a ticket is delayed, if a billing dispute affects good standing, if a board election is challenged, or if a court dispute freezes the registry's bank account, the operator has no realistic alternative registry to which it can move the same African number resources. It cannot take those prefixes to ARIN, RIPE NCC, APNIC or LACNIC as a dissatisfied buyer might change software vendors, auditors or trade associations.
That practical captivity is the starting point for any serious discussion of AFRINIC membership accountability. Membership in a regional internet registry is not ordinary club membership. It is not membership in a professional body that an operator can drop without touching its balance sheet. It is closer to compulsory participation in a monopoly ledger. AFRINIC records who is recognised for scarce internet number resources in Africa and parts of the Indian Ocean. It also provides, or controls access to, the services that turn that recognition into operational continuity: WHOIS and RDAP publication, reverse DNS delegation, resource certification, routing-registry functions, billing status, transfer processing, resource-request handling, dispute channels and administrative records that make a network visible as a legitimate holder.
The economic mechanism follows from the monopoly. Where exit is absent, voice and accountability have to carry more weight. Accountability is not a courtesy owed by a nonprofit to its subscribers. It is the discipline that protects captive members from uncontrolled fees, opaque legal spending, weak elections, discretionary enforcement, board capture and institutional self-preservation. A registry board can influence how much members pay, which disputes are fought or settled, how scarce IPv4 policy is interpreted, how aggressively the registry reviews resource use, how much information members receive, whether management is disciplined, whether temporary receivership practices become normal, and whether a contested election is treated as a minor irregularity or as a threat to the authority of every subsequent corporate act.
AFRINIC is the test case because its crisis made those questions concrete. Public reporting and court-facing statements have described a long sequence of stress: reported address-record corruption and a former senior staff member's alleged role in the movement or sale of valuable IPv4 space; AFRINIC's dispute with Cloud Innovation over use of a large IPv4 holding and the possibility of withdrawal; litigation that contributed to frozen bank accounts and institutional paralysis; years without a functioning board or ordinary chief executive appointment; receivership ordered by the Mauritian courts to preserve the organisation and arrange elections; a June 2025 election suspended and then annulled after concerns over powers of attorney and voter authority; a later election that restored a formal board without eliminating the litigation overhang; and renewed 2026 fights over winding-up proceedings, bylaws, member rights, transfer policy and statements about commercial IPv4 leasing.
Those facts matter, but the article is not another general essay on registry-layer risk. The question here is narrower and more institutional: who can vote, who receives information, who controls budgets, who decides litigation posture, how members discipline directors and management, why election legitimacy and voter records matter, and why court-appointed continuity cannot substitute for accountable member governance. AFRINIC's crisis shows that a registry can keep enough technical services alive while still failing the membership bargain. Protecting the registry function is not the same as protecting every incumbent institutional claim. The ledger must continue, and members must control its operator.
That distinction is central. In a trade body, a failed election may embarrass the board. In a monopoly registry, it can weaken confidence in who may set budgets, appoint executives, supervise scarce-resource policy and speak for the recognised holder base. In a voluntary association, opaque legal spending may annoy members. In a captive ledger, it can consume fees paid by operators who require service continuity and cannot relocate their records. In a normal club, a messy membership roll may be a procedural defect. In AFRINIC's circumstances, it touches the authority of the institution that records scarce, tradeable, financeable and operationally necessary identifiers.
The institutional-economics lesson is therefore stark. A monopoly ledger cannot rely on the soft legitimacy of community language. It must earn legitimacy through hard controls: a defensible member register, clean elections, usable information rights, budget visibility, litigation discipline, narrow mandate boundaries and a continuity firewall that keeps the registry function alive without converting emergency administration into permanent consent.
Membership is captive participation, not association life
The word member softens the problem. It suggests consent, affiliation and shared purpose. In the RIR system, membership also describes dependency. A resource member joins because it needs number resources, or because it already holds them and must maintain the relationship through which the registry recognises and services those records. AFRINIC's fee model makes the relationship more than symbolic. Local Internet Registry members pay annual fees by resource category. End-site organisations pay under a different schedule. New allocations or assignments can carry separate charges. Good standing can affect transfers, reverse DNS changes, access to services and continued administrative recognition.
For a large carrier, the annual registry fee may be small compared with network revenue. For a small African operator, it may be material. Yet the visible fee is not the larger cost. The larger cost is dependence on a single recognised recordkeeper. A small ISP needs registry records accurate enough for peers and counterparties. A data-centre operator may need reverse DNS to satisfy customers. A regional network may need RPKI publication to reduce route-origin risk. A buyer or lessee of IPv4 capacity needs the recognised holder position to be clear enough for contracts. A lender looking at network expansion needs to know whether the address layer can be relied upon. In each case, AFRINIC membership is the administrative doorway through which operational facts become public recognition.
Ordinary associations can tolerate more looseness because market discipline is available. If a chamber of commerce becomes wasteful, a company can resign and continue trading. If an industry conference mishandles elections, sponsors can move to another event. If a voluntary standards group loses authority, implementers may still use the underlying technology while ignoring the organisation. A regional internet registry is different. Its members cannot easily replicate the ledger, and the rest of the internet coordinates around the recognised registry system. Even critics who argue for restructuring understand the point: current registry recognition is not just another service. It is embedded in how the market verifies resource status.
The absence of exit turns ordinary governance defects into economic hazards. A poor board decision can raise fees for members who cannot avoid the fee base. A weak dispute process can trap resources in uncertainty. A board captured by a faction can steer legal strategy toward institutional self-protection or toward one commercial camp's interests. A deficient voter register can make every subsequent reform suspect. A lack of transparent legal budgets can force members to fund court battles whose objectives, costs and settlement thresholds they do not understand. A vague mandate can let the registry move from recordkeeping into discretionary control over commercial conduct while leaving the affected businesses to absorb the economic loss.
This is the member-control problem at the centre of AFRINIC. Members are not only stakeholders in an association. They are users of a settlement utility. Their payments keep the utility alive. Their records give the utility its practical importance. Their votes are supposed to discipline the directors who supervise the utility. Their information rights are supposed to let them judge whether money is being spent on service continuity, institution-building, litigation, staff, training, policy administration or defensive campaigns. If the voting base is unclear, the board is disputed, the receiver is temporary and the courts are overloaded with governance questions, the link between payment and control breaks.
The break matters more under scarcity. IPv4 addresses remain valuable because IPv6 deployment has not removed demand for IPv4 compatibility. AFRINIC has long administered a region that retained available IPv4 space after several other RIRs were effectively exhausted. Its own exhaustion materials describe soft-landing phases, final-pool constraints, resource-request queues, utilisation requirements and maximum allocation sizes. Public reporting in 2026 described a remaining pool large enough to matter to operators but small enough to make allocation policy politically sensitive. Scarcity makes registry procedure economically sensitive. The power to approve, delay, deny, reserve, recognise, transfer, publish, certify or challenge a resource position affects value.
That is why the accountability standard for AFRINIC should be higher than the standard for an ordinary nonprofit. Members should not have to infer from scattered statements whether the registry is spending their fees prudently, whether a lawsuit threatens service continuity, whether a board's authority is clean, whether a voter list is reliable, or whether resource policy is being used as a disguised capital-control tool. Nor should they be forced to treat every official assurance as sufficient. A captive monopoly ledger earns trust by making its constraints, incentives and decision rights visible.
The practical implication is uncomfortable for both defenders and critics of the institution. Defenders cannot answer every challenge by invoking continuity or the African internet community. Critics cannot answer every registry action by asserting property rights or commercial freedom. The real question is whether the governance design matches the dependency. A compulsory ledger needs a member-control architecture strong enough to absorb disagreement without turning every disagreement into a corporate emergency.
The corruption story was also a membership story
AFRINIC's recent legal crisis did not arise in a vacuum. Years before the board paralysis and the Cloud Innovation dispute became the dominant public narrative, the registry had to confront reported address-record corruption. In 2019, investigative reporting described allegations that a senior AFRINIC staff figure had been linked to companies involved in the sale or movement of African IPv4 address space. The reporting, based partly on the work of researcher Ron Guilmette and journalists in South Africa, placed the alleged value of affected address blocks in the tens of millions of dollars. AFRINIC's then leadership acknowledged awareness of the allegations and an investigation. The staff figure resigned.
The details are important, but the institutional lesson is broader. A registry's membership model depends on members believing that records are not quietly altered, misused, reassigned or commercialised through privileged access. If the public ledger can be manipulated by insiders, the harm is not confined to the immediate victim. Every member faces a higher diligence burden. Historical records become suspect. Dormant allocations become targets. Resource holders ask whether old company changes, acquisitions, closures and contact data were properly captured. Buyers and lessors ask whether a block's registry history is clean. Operators ask whether staff access controls are strong enough. The registry's public purpose is weakened because record integrity is the foundation of membership trust.
Corruption history also changes how members interpret later enforcement. When a registry with a clean record enforces rules, members may disagree but still assume institutional neutrality. When a registry has a public history of record-integrity failure, enforcement looks different. A member targeted for review can claim selective treatment. Other members can wonder whether the same standards are applied consistently. Critics can argue that the registry should repair its own controls before making expansive claims over commercial use. Defenders can reply that stronger enforcement is exactly what is needed after historical abuse. Both positions can be plausible. That is why accountability cannot depend on trust alone.
The member-economics question is precise: what controls allow the fee-paying and resource-dependent membership to know whether the registry learned from the record failure? A credible regime would give members usable answers on audit practices, privileged access, historical remediation, whistleblowing outcomes, conflict disclosures, escalation procedures, publication of material control failures and independent review of disputed changes. It would distinguish public registry integrity from reputational messaging. It would be clear about which categories of record were affected, what risks remain and how future misuse would be detected.
AFRINIC's later crisis made that precision more necessary, not less. Once litigation placed the organisation under financial and governance strain, members had to rely even more heavily on staff, temporary officeholders and court-supervised arrangements. A receiver can preserve assets and keep operations going, but the daily work of a registry still runs through people, credentials, databases, tickets, approvals and internal controls. If the board is absent, the chief executive role is not normalised and major disputes are live, the control environment becomes a matter of member protection.
The corruption story also demonstrates why official mission language is insufficient. AFRINIC describes itself as a nonprofit, member-based organisation serving Africa and the Indian Ocean region, entrusted with number resources and committed to efficient service. Those statements define the role. They do not prove that the role is being exercised accountably. A monopoly ledger cannot satisfy members merely by saying it serves the community. It must show that the mechanisms by which records are changed, resources are reviewed, fees are spent and disputes are escalated are resilient against insiders, factions and external pressure.
For a small operator, this is not an abstract ethics question. If a registry once suffered alleged insider misuse of records and later enters a period of boardlessness and receivership, the rational response is to ask for more visibility. Who can alter its records? What review exists if a change is made in error? How are dormant resources protected? How are staff conflicts checked? How are allegations reported without compromising investigations? How much of the budget goes to control remediation rather than public relations or endless litigation? Membership accountability is the channel through which those questions become institutional duties rather than private worries.
That is why the old corruption story belongs inside the membership story. It is not only evidence that a past control failed. It is evidence that registry membership depends on something stronger than confidence in the good intentions of staff and directors. Members fund the ledger because they must. They are entitled to know that the ledger is harder to corrupt, easier to audit and more accountable to them than it was before the failure became public.
Cloud Innovation turned enforcement into a control question
The dispute between AFRINIC and Cloud Innovation is often described through competing claims about IPv4 use, contractual rights, regional purpose and commercialisation. That description is accurate but incomplete. The dispute also turned enforcement into a question of who controls the registry, who pays for that control, and how much discretion a registry may exercise over scarce resources once a member's business depends on them.
Public reporting has described AFRINIC's 2021 position as an allegation that Cloud Innovation had breached its service agreement by using IPv4 resources inappropriately, with procedures that could have led to withdrawal of resources. Cloud Innovation treated the matter as existential and brought multiple legal actions. Those actions, and related court orders, contributed to bank-account freezes and institutional paralysis. AFRINIC was left unable for a period to appoint a board or chief executive and unable to perform some normal functions. Later AFRINIC statements characterised continuing litigation and procedural challenges as attempts to paralyse the organisation. Cloud Innovation and associated critics responded that the deeper problem was structural: a registry designed as an administrative recordkeeper had come to exercise high-consequence power over economically critical resources without equivalent liability.
One need not accept either side's whole narrative to see the accountability failure. A resource-withdrawal dispute involving a large IPv4 holder should be governed by a process whose boundaries, evidence standards, appeal channels, liability consequences and member oversight are clear before the fight begins. If the registry can threaten withdrawal from a scarce-resource holder, the holder will naturally seek every legal tool available. If the holder can immobilise the registry through litigation, other members will naturally ask why their fees and services are exposed to one dispute. If the board and management cannot absorb the conflict, members will ask whether enforcement discretion exceeded institutional capacity.
The problem is not only whether Cloud Innovation was right or AFRINIC was right. The problem is that the system lacked a durable way to keep a member enforcement dispute from becoming an institutional-continuity crisis. A monopoly ledger should be able to enforce registration rules, contractual obligations and policy requirements without turning each high-value dispute into a struggle over the registry itself. It should also be able to protect uninvolved members from the blast radius of a dispute they did not create.
That is why membership accountability cannot be reduced to annual votes. Members needed to know much earlier what enforcement posture the registry was taking, what legal risk accompanied that posture, what costs were being incurred, how the board assessed the impact on services, whether settlement options existed, whether an independent forum could narrow the issues, and how the rights of uninvolved members would be protected if litigation escalated. They also needed to know whether the registry's interpretation of its mandate was changing. Was AFRINIC acting as a narrow ledger enforcing registration accuracy and policy compliance, or as a broader gatekeeper trying to police the commercial use of IPv4 resources? The answer affects every member whose resources have market value.
The board-control element follows. When a dispute concerns scarce assets worth large sums, control of the board becomes economically significant. The board can select executives, approve budgets, direct litigation, authorise settlements, oversee bylaws, shape committee appointments, decide how much information members receive and influence how policy language is implemented. In a low-stakes association, a board election is mostly a governance exercise. In AFRINIC's setting, a board election can determine the future treatment of valuable IPv4 holdings, interregional transfer possibilities, leasing models, fee levels, legal spending and institutional alliances.
That does not mean every candidate backed by a commercial critic is illegitimate, or that every candidate backed by establishment groups is captured. It means member accountability must be designed for a world in which candidates, proxies, credentials and voting authority are economically valuable. A registry election is not safe merely because the bylaws contain election provisions. It needs hardened identity verification, auditable appointment of representatives, limits and disclosure around proxies or powers of attorney, independent observation, timely incident reporting, clear challenge procedures and a member register that is not a moving target.
The dispute also shows the danger of legal-budget incentives. A registry under attack may spend heavily to defend itself, and sometimes it must. But if legal spending is funded by captive members, and if the board or temporary leadership has limited accountability to those members, litigation can become self-justifying. The institution fights because its authority is challenged. The challenger continues because that authority is the disputed asset. Members pay because they need registry services. Without transparent budget discipline, legal strategy can consume the utility it claims to protect.
The narrow-ledger principle offers a way out. AFRINIC should be powerful where the ledger requires power: accurate records, clear eligibility, fair process, security publication, reverse DNS, transfer recording and dispute handling within defined limits. It should be restrained where enforcement becomes economic planning. The more a registry decision affects a member's commercial use of scarce IPv4 resources, the more process, review and member visibility are required. That is not a concession to any particular litigant. It is the price of legitimate monopoly administration.
Voting legitimacy has market consequences
The June 2025 AFRINIC election was supposed to move the organisation from receivership toward normal governance. Instead, it became another exhibit in the economics of membership accountability. Electronic voting began after legal challenges and concerns over the nominations process. In-person voting followed on June 23. Minutes before the voting period ended, the nominations committee suspended the process, citing questions about powers of attorney and voting authority. The receiver later annulled the election, referring to stakeholder concerns and potential irregularities in voter documentation. Public reporting described allegations that some representatives discovered votes or powers of attorney had been lodged on behalf of members without proper consent, and that one party was said to hold powers of attorney for a large share of the resource-holder base. ICANN demanded explanations and raised the possibility of compliance review. Cloud Innovation argued that annulment over a disputed vote or proxy could trap the organisation in endless contests.
The precise facts of every alleged power-of-attorney issue are for courts, investigators and participants to test. The economic point is already clear. In a captive monopoly ledger, voting legitimacy is part of the control system. It determines who may discipline the board, approve or resist bylaws, judge legal spending, change fee structures, select institutional priorities and restore confidence after crisis. If members do not know whether votes were cast by authorised people, the board that emerges cannot easily command trust, especially when major litigation and resource-policy disputes remain alive.
This is not a sentimental point about participation. It is a question of transaction costs. A clean election lowers the cost of relying on the institution. Members, counterparties, courts, banks, operators and global coordination bodies can treat board decisions as decisions of a legitimate corporate organ. A disputed election raises the cost of every subsequent act. Was the board properly seated? Was the budget validly approved? Was the bylaw committee properly constituted? Was the litigation posture authorised? Were staff appointments and service changes made under clean authority? The more valuable the registry's records, the more those questions matter.
AFRINIC's later 2025 election, which produced eight directors and allowed the board to convene for the first time since 2022, reduced one immediate failure: the absence of formal directors. It did not erase the accountability question. Public reporting noted that critics were likely to ask Mauritian courts whether the election arrangements were consistent with the organisation's rules. It also noted discomfort among some community actors over the influence of external continental bodies and the majority support certain candidates received from Smart Africa. Such concerns may or may not prove legally decisive. Their importance lies in showing that, after a crisis, legitimacy is not restored by counting seats alone. The process by which the seats are filled must be strong enough that losing factions, courts and ordinary members can distinguish defeat from disenfranchisement.
The June annulment also exposed an information problem. Members were asked to trust that annulment was necessary, yet public reporting described limited detail from AFRINIC and the receiver. ICANN demanded fuller reporting. Members who had paid fees and depended on the registry were left with uncertainty about whether the process failed because of widespread fraud, isolated irregularity, poor documentation rules, inadequate verification, overbroad proxy arrangements, factional pressure or excessive risk aversion by the receiver. Those are very different diagnoses. Each implies a different repair. Without adequate disclosure, members cannot know whether the next election is safer.
An accountable membership system would treat election infrastructure as a registry service, not a meeting formality. The member register would be reconciled in advance. The distinction between resource members, registered members under Mauritian company law, directors and voting representatives would be explained in plain language. Proxy and power-of-attorney rules would be narrow, verified and consistently applied across online and in-person channels. Any person claiming authority to vote for another member would be validated against records controlled by the member, not merely against documents presented at the meeting. Incident reporting would be timely and specific enough to reassure honest participants without prejudicing investigations. Annulment would trigger an automatic independent report.
The reason is simple. Members are captive, so the election is the principal mechanism through which they convert payment into control. If that mechanism fails, the registry's monopoly becomes harder to justify. If voting authority can be aggregated, manufactured, confused or challenged after the fact, then control of the ledger becomes a prize rather than a mandate. That is why voter records are not administrative trivia. They are part of the economic security of the registry.
The registered-member gap goes to the heart of the model
AFRINIC's 2026 bylaws debate brought a deeper legal and economic problem into public view: the possible dissonance between resource membership in the registry sense and registered membership under Mauritian company law. Public reporting described an ISPA position that many rights provided to AFRINIC resource members under the bylaws could be legally challenged because Mauritian company law limits certain governance functions to registered members. ISPA's proposed approach, as reported, was to clarify that only directors are registered members while creating community-resolution mechanisms for resource-member participation. Critics, including the Number Resource Society, attacked the idea as a reduction in real member rights.
This is not a technicality. It goes to the legitimacy of the membership bargain. AFRINIC's users pay as resource members and depend on the registry's services. The institution's public identity is member-based. Yet if the legal structure beneath that identity leaves resource members with rights that are vulnerable to challenge, the accountability system is weaker than advertised. Members may believe they are the governing constituency, while corporate law may treat the formal company differently. A crisis then turns the gap into litigation fuel.
The June 2025 court and election controversy included a related issue. Cloud Innovation was reportedly listed as a registered member or shareholder in Mauritian corporate records, a classification that public reporting said the court later treated as erroneous and attributable to the registrar rather than AFRINIC or the receiver. The court required communication to members clarifying the matter, while ICANN and others remained concerned about election integrity. The episode showed how easily corporate classification, resource membership, voting rights and control narratives can become entangled. In a high-value registry, even a temporary or erroneous entry can become a claim, a fear or a strategic exhibit.
An ordinary association might survive ambiguity between subscriber categories and formal company membership. A monopoly ledger should not. The more captive the users, the more explicit the legal rights must be. If resource members fund the institution and are subject to its resource policies, their governance rights should be defined in a form that courts can enforce and members can understand. If some governance functions must legally reside with registered members or directors, that limitation should not be hidden behind community language. If community resolutions are used, their legal effect should be explicit. If directors are the formal members, resource members need other enforceable controls over budget, elections, litigation reporting, service standards, board recall, conflicts and bylaw changes.
The economic question is not whether every resource member should have identical voting rights on every corporate matter. It is whether the governance design matches the dependency. A member with a small ASN-only or end-site account may not need the same exposure to complex corporate decisions as a large LIR, but it still needs protection against arbitrary registry action, fee misuse and loss of service continuity. A large holder may have more financial exposure, but it should not be able to buy control through proxies or legal pressure. A government-linked or continental-policy body may have legitimate regional-development interests, but it should not convert those interests into control over a private ledger without accountable member consent. A registry's bylaws must allocate these tensions openly.
The registered-member gap also illustrates why courts cannot be the normal accountability mechanism. Courts can interpret company law, correct erroneous records, supervise receivership, restrain false statements, admit interveners and decide winding-up applications. They cannot by themselves create the economic legitimacy that comes from a clearly understood membership contract. If every major governance question must be litigated, the model has already failed to provide low-cost coordination.
AFRINIC's path forward therefore depends on more than seating a board. It requires a legal architecture in which the resource-dependent membership knows exactly what it controls, what it can inspect, how it can challenge directors, how it can vote, how representatives are verified, how bylaws are changed and where registry discretion ends. Without that clarity, the phrase member-based remains too weak for a compulsory registry relationship.
Receivership is emergency plumbing, not consent
Receivership is often described as proof of resilience. In one sense, it was. The Mauritian court's appointment of an official receiver in 2023 was intended to preserve AFRINIC's assets, prevent relocation or takeover, maintain the status quo, oversee elections, facilitate a proper board and support appointment of executive leadership. The NRO welcomed the development as a path toward functional governance and continued registry services. Public commentary argued that receivership showed rule-of-law safeguards working: the organisation could be stabilised under court supervision rather than simply collapse.
That conclusion is partly right and economically incomplete. Receivership can preserve continuity. It cannot substitute for accountable membership governance. A receiver is emergency plumbing, not a durable consent mechanism. The receiver's task is to maintain the business, organise the conditions for governance and comply with court directions. The receiver does not become the membership. The receiver does not resolve every legitimacy question. The receiver does not give ordinary resource holders control over legal budgets, enforcement philosophy, transfer policy, bylaw reform or board priorities. The receiver's authority comes from court order, not from the ongoing bargaining relationship between the registry and its captive users.
AFRINIC demonstrates both sides of the distinction. During receivership, services did not simply vanish. Staff continued enough of the registry function to prevent immediate systemic failure. That matters. It shows the value of a continuity firewall: the ledger, technical services and ordinary member support should keep running even when the corporate body is in distress. But the same period also produced delayed governance restoration, questions about the receiver and nominations committee, postponed elections, annulment, demands for explanation, and a second election whose legitimacy critics continued to scrutinise. The receiver could keep the vessel afloat; he could not make the membership accept that every governance choice was legitimate.
Continuity can therefore become a misleading standard. If the internet keeps routing, defenders of the incumbent system may say the crisis was contained. If the registry website remains online and tickets are answered, they may say the model worked. But members experience costs long before catastrophic outage. Transfers may be delayed. New allocations may be constrained. Staff attention may be consumed by legal matters. Member confidence may fall. Counterparties may demand extra representations. Budget planning may become uncertain. Policy development may stall. The board may be absent. Litigation may consume funds. Elections may become battlegrounds. These are real economic costs even if packets continue to flow.
Receivership also changes incentives. A receiver must be cautious, especially when litigation is live and stakeholders are hostile. Caution may protect the estate, but it may also slow decisions that members need. A receiver may rely on external advisers or committees to reduce risk, increasing cost and complexity. A receiver may avoid taking substantive positions, leaving disputes unresolved. A receiver may annul an election to protect legitimacy, but without a full explanation the annulment itself becomes a legitimacy problem. These are not moral criticisms. They are structural limitations of crisis administration.
The proper lesson is that a registry needs pre-crisis accountability strong enough that receivership is rare and narrow. Essential registry services should be separated from board disputes, litigation spikes and corporate-control battles. Bank accounts for ordinary operations should be protected from total immobilisation where law permits. Technical publication functions should have documented service-continuity plans. Member support should be insulated from legal strategy. Election infrastructure should be hardened before a vote is called. Emergency governance should be defined before a crisis, not improvised under court pressure.
That firewall must not become an incumbency shield. Preserving the registry function is different from preserving every authority claim made by the current institution. If members are told that any challenge to the board, bylaws, legal spending or enforcement posture endangers continuity, then continuity becomes a device for suppressing accountability. AFRINIC's crisis shows the opposite need: protect the ledger so members can challenge leadership without threatening the records on which their businesses depend.
Courts are indispensable when a corporate registry breaks down. They are not a replacement for member control. The goal is not to keep AFRINIC permanently under judicial rescue. It is to make the membership architecture strong enough that courts handle exceptional disputes while members govern the ordinary economics of the monopoly they fund.
Legal spending reveals who really controls the institution
The most concrete test of a member-funded institution is how it spends money when threatened. AFRINIC's public fee model funds operations, member services, policy administration, technical services, staff, training, outreach and governance. In an ordinary year, members may debate whether those functions are efficient. In a crisis year, the budget can be pulled toward lawyers, court filings, external advisers, election contractors, communications, forensic review and defensive institutional activity. The question is not whether legal spending is ever justified. It often is. The question is whether captive members can see enough to judge proportionality.
AFRINIC's dispute environment makes that unavoidable. Public reporting has referred to more than twenty lawsuits or a complex set of interlocking matters. AFRINIC has accused Cloud Innovation, Larus and associated campaigns of creating procedural roadblocks and a web of litigation. Cloud Innovation and its allies have argued that litigation is a response to concentrated registry power. ICANN has intervened at different points, including in relation to elections and a winding-up application. Mauritian courts have been asked to consider receivership, corporate records, election arrangements, takedown statements and the status of numbering resources in liquidation. Each step carries cost.
For members, legal budgets are not neutral. Money spent on litigation is money not spent on registry resilience, staff capacity, service improvement, record audits, RPKI reliability, member support, training, security, data-quality repair or fee reduction. Litigation may be necessary to protect those same functions, but necessity cannot be asserted indefinitely without becoming a blank cheque. A monopoly ledger funded by captive users needs a discipline that separates defensive legal spending from institutional self-preservation, principled mandate defence from factional strategy, and unavoidable compliance costs from avoidable escalation.
The board's role is central. Directors control, or should control, the legal posture. They decide whether to fight, settle, appeal, disclose, insure, reserve funds, seek member approval, publish summaries, commission independent review, or ring-fence operations. If the board is absent, disputed, newly seated under challenge, or dependent on receiver-era arrangements, legal strategy becomes harder for members to evaluate. The fee-paying base may be told that legal action is necessary, but not what success looks like, what the budget envelope is, how risk is measured, or what settlement would cost compared with continued litigation.
This is a principal problem, but with higher stakes because exit is weak. Members delegate authority to directors and administrators. Directors and administrators may rationally prefer to defend institutional authority, avoid personal blame, preserve relationships with global coordination bodies, or resist a litigant whose conduct they view as harmful. Those preferences may align with member interests. They may also overshoot. Without reporting, members cannot tell. The result is distrust: critics assume legal spending protects the gatekeeper, defenders assume every challenge is sabotage, and ordinary operators pay and wait.
An accountable registry would publish a member-facing litigation framework. It would not disclose privileged strategy, but it would disclose categories of proceedings, aggregate legal spend, budget variance, material service impacts, reserve assumptions, insurance position, board approvals, settlement principles and the decision criteria for continuing or resolving major disputes. It would distinguish cases that threaten registry continuity from cases that concern ordinary contractual enforcement. It would explain when a dispute could affect fees or services. It would give members periodic, plain-language reports without turning legal updates into public campaigning.
AFRINIC's own public comments about rising legal costs and delayed community initiatives show why this matters. If litigation obstructs training, research and membership strengthening, members should know the scale of that diversion. If certain lawsuits seek to block IPv4 issuance or bylaw review, members should know how the board is balancing compliance with service restoration. If a winding-up application threatens the corporate shell, members should know what continuity plans protect the ledger and what legal theory protects numbering resources from being treated as distributable corporate assets. If a takedown order concerns public statements about leasing and court approval, members should know whether the issue affects ordinary members' commercial arrangements or only particular representations.
The point is not to litigate by referendum. It is to prevent legal strategy from becoming the area where the monopoly is least accountable precisely when member exposure is greatest. If members fund the defence of the ledger, they need visibility into the defence. If they are asked to bear exceptional legal cost, they need to know whether the cost protects service continuity, institutional pride, board discretion or one side's view of the address market.
Scarcity makes governance a market issue
IPv4 scarcity changes the meaning of governance. When a registry mainly distributes abundant resources, board control is important but less likely to move capital values. When the resource is scarce, board control affects not only institutional reputation but also the value, liquidity and risk profile of existing holdings. AFRINIC is in the scarce-resource phase. Its public exhaustion materials describe soft-landing rules, limited-pool management, justified-need review, maximum allocation sizes, utilisation thresholds and final-pool constraints. Public reporting in 2026 described a remaining unallocated pool of 773,376 IPv4 addresses, a number large enough to matter to operators but small enough to make allocation policy and timing politically salient.
Scarcity gives several board powers market significance. The board can oversee implementation of transfer policies. It can ratify or reject policy outputs where the process requires board action. It can influence staff interpretation of service agreements and compliance checks. It can approve fees that change the carrying cost of address holdings. It can supervise how aggressively the registry treats leasing, out-of-region use, dormant records or address-market intermediaries. It can direct litigation against large holders or decide whether to settle. It can shape bylaw reforms that determine who gets to vote on future rules. It can decide how much transparency members receive about remaining inventory and allocation queues.
These powers are not ownership in the ordinary sense. AFRINIC and many in the RIR system maintain that internet number resources are not owned as traditional property, and policy documents often describe holders as custodians rather than owners. There are good reasons for caution. IP addresses are part of a global coordination system. Their uniqueness depends on shared recognition. Routing depends on operational acceptance. A registry record is not identical to land title or a share certificate. Treating number resources as ordinary property could create its own legal and coordination problems.
Yet avoiding ownership language does not avoid economics. IPv4 blocks are bought, sold, leased, financed, valued in corporate transactions and used as inputs into customer revenue. Companies plan around them. Courts are asked to consider their treatment. Banks and buyers ask whether a holder's position is secure. Members pay fees to maintain recognition. A registry decision that withdraws, freezes, delays, challenges or restricts recognition can produce real losses even if the legal theory says the member never owned the resource in the traditional sense.
This is the asset-recognition problem. A resource does not have to be full property to be economically important. Many valuable positions are conditional: licences, concessions, spectrum rights, payment-system access, clearing memberships, permits, leaseholds and contractual entitlements. The more valuable and non-substitutable the position, the stronger the accountability required of the body that administers recognition. AFRINIC membership should be understood in that class. Members may not own number resources as freehold property, but they hold recognised positions that support business value and operational continuity.
The registry must therefore be careful about discretionary power. If AFRINIC asserts that resources are not property and also claims broad authority to determine acceptable use, transferability, leasing implications or regional economic purpose, members face a one-sided risk. Their reliance is economically real, but the registry's liability may be limited. Their exit is impractical, but the registry's mandate may be broad. Their fees fund the institution, but the institution may spend on litigation that protects its own interpretation of power. That structure demands stronger member oversight.
The Cloud Innovation dispute is the most visible example, but the principle affects smaller members too. A small ISP may never lease addresses globally or litigate over a large block. It still depends on asset recognition in a narrower sense. Its addresses support customer contracts. Its ASN supports routing identity. Its reverse DNS may support mail deliverability and enterprise services. Its registry status can affect credibility with peers and upstreams. If the registry's governance fails, the small member's resource position may be discounted by counterparties even though it did nothing wrong.
Asset recognition also changes fee politics. Members do not pay merely for staff time. They pay for a recognition environment that preserves the value of their operational positions. If legal disputes, poor elections or opaque controls raise uncertainty, members are effectively paying for a lower-quality ledger. Conversely, if the registry invests in auditability, service reliability, independent dispute resolution and clear member rights, it enhances the value of member positions. Accountability is not a cost centre; it is part of the asset-recognition service.
The danger is that both sides in registry disputes can misuse asset language. Holders can overstate property claims and ignore the coordination function. Registries can deny property language and understate economic harm. AFRINIC's lesson is to separate the two. The registry need not concede that IPv4 addresses are ordinary property in order to accept that recognised resource positions have economic value requiring due process, transparency, liability awareness and member control. Holders need not deny community coordination in order to demand predictable and accountable administration.
Scarcity also raises the cost of bad elections. If the board controls only meeting schedules, a flawed vote is annoying. If the board can influence the future of scarce IPv4 value, a flawed vote becomes a market event. Proxy rules, voter eligibility, nominations committees and member-register accuracy are therefore part of the capital structure around number resources. This is why allegations around powers of attorney in AFRINIC's June 2025 election were so damaging. They suggested that voting authority itself might be a scarce asset worth capturing.
The more valuable IPv4 becomes, the less plausible it is to govern the registry like a small technical association. AFRINIC's board must be accountable because it sits between a monopoly ledger and a scarcity market. That position is too economically powerful to rest on trust, custom or emergency repair.
The mandate firewall: ledger first, community second
AFRINIC's public mission includes service to the African internet community, internet self-governance, training, development and efficient distribution of resources. Those objectives are not illegitimate. In a region with uneven connectivity, limited technical capacity in some markets and real development needs, training and community support can matter. The problem arises when community language expands the registry's mandate beyond the narrow ledger functions that justify compulsory membership.
The narrow ledger is the reason operators must deal with AFRINIC. It records and services unique number-resource claims. It supports registration accuracy, reverse DNS, public data, routing-security publication, transfer records, resource eligibility and coordination with the global system. These functions lower transaction costs because they are predictable and limited. The more the registry moves into judging the economic desirability of members' business models, the more it becomes a gatekeeper rather than a settlement utility.
AFRINIC's crisis shows how quickly the line blurs. The registry may believe that restricting certain transfers protects Africa's resource base. Critics may argue that such restrictions trap capital and punish holders for using resources commercially. The registry may believe that action against a large holder enforces the service agreement. The holder may argue that the registry is trying to destroy asset value without adequate liability. The registry may believe that coordinated continental support protects institutional continuity. Some members may fear political or governmental capture. The registry may believe that litigation threatens community programmes. Critics may say those programmes are being used to shield discretionary power.
The mandate firewall does not decide every policy question in advance. It requires that broad social or regional objectives be separated from core ledger integrity. If a policy limits transfers, the economic rationale, legal basis, member mandate and expected costs should be explicit. If AFRINIC investigates resource use, the evidence standard and appeal channel should be clear. If the registry promotes IPv6, it should not use that promotion to obscure the continuing value of IPv4 records. If the board spends on development programmes, members should know how those programmes are funded relative to essential registry operations and legal costs. If the institution claims to protect the African community, it should specify which accountable members authorised the action and how dissenting members can challenge it.
This distinction is especially important because AFRINIC is the only RIR for its region. Regional uniqueness gives the institution a public-sounding role, but it does not eliminate the private-law and member-funded character of the organisation. A company in Mauritius, a data centre in Kenya, a network in Nigeria, a university in South Africa and an operator in the Indian Ocean may all depend on the same registry. Their interests are not identical. The phrase African internet community can conceal conflicts among incumbents, new entrants, governments, academic networks, carriers, address-market participants, civil-society groups and global coordination bodies. A board that claims to speak for all of them needs stronger accountability, not weaker scrutiny.
The mandate firewall also changes how official RIR, NRO, ICANN and AFRINIC statements should be used. They are useful exhibits about actions taken, dates, institutional positions and formal concerns. They are not by themselves a sufficient explanation of member economics. ICANN may warn about election integrity or seek standing in winding-up proceedings. The NRO may welcome receivership. AFRINIC may state that a litigant is trying to paralyse it. Cloud Innovation may argue that the registry model is structurally broken. Each statement is evidence of a position. Members still need independent disclosure from the institution they fund: what is the cost, what is the risk, what service is affected, what decision is required, and what control do members have?
The firewall also protects AFRINIC itself. A narrow ledger is easier to defend. Courts, members and counterparties can understand why continuity matters: the registry keeps unique identifiers recognised and operational. A broad gatekeeper is harder to defend because every enforcement decision looks like economic planning. When the registry's mandate is open-ended, opponents can attack it as arbitrary; defenders can justify almost anything as community protection. The result is a legitimacy fight in which technical services become collateral.
Membership accountability is how the firewall is maintained. Members should approve the budgetary weight given to community programmes. They should see how policy proposals connect to ledger integrity rather than factional advantage. They should be able to challenge mandate expansion. They should know whether directors have interests in address markets, government programmes, vendors, advocacy bodies or competing institutions. They should receive legal explanations when corporate law limits their formal power. Without those controls, the registry's monopoly status becomes a licence to define its own purpose.
AFRINIC's next phase will test whether it can become narrower without becoming weaker. A narrow registry is not a passive registry. It enforces accuracy, security, eligibility and fair process. But it does so within limits that members can inspect and courts can recognise. That is the institutional discipline a scarce-resource ledger needs.
Information rights are the cheapest form of insurance
Many registry failures become expensive because members learn too little too late. AFRINIC's crisis contains repeated examples of information arriving after trust had already weakened: public allegations of address-record corruption; surprise and confusion over bank-account freezes and litigation effects; limited visibility into the reasons for election annulment; uncertainty over powers of attorney; conflicting claims about corporate membership status; concern over whether newly elected directors were vulnerable to court challenge; public disputes over bylaws and resource-member rights; and continued disagreement over whether litigation, transfer policy or leasing statements threatened continuity.
Information rights are not a bureaucratic luxury in that environment. They are a low-cost insurance mechanism. Members who receive timely, usable information can price risk, plan operations, participate intelligently and discipline leadership before disputes become existential. Members kept in the dark respond through rumours, factional mailing lists, litigation, public accusations, defensive contracts and distrust of official communications. The information deficit creates the instability the institution then cites as a reason for centralised control.
The information that matters is not endless publication of documents. It is structured, decision-useful disclosure. Members need service-continuity reports: ticket performance, registry-system uptime, RPKI and reverse DNS status, data-quality remediation, staffing constraints and known service risks. They need financial reports that distinguish ordinary operations, community programmes, capital investment, legal expense and exceptional recovery costs. They need governance reports that explain board decisions, conflicts, committee appointments, bylaw proposals and election readiness. They need litigation summaries that identify material disputes without waiving privilege. They need membership-register reports that explain categories, voting eligibility, representative verification and changes affecting member rights.
AFRINIC's public materials already show the categories of activity that require such disclosure. It runs member services, number-resource management, reverse DNS, WHOIS, RDAP, IRR and RPKI. It charges fees and applies late-payment steps. It evaluates resource requests under policy. It runs elections and committees. It supports training and development activities. It exists under Mauritian corporate law while participating in a global RIR system. Each category creates an accountability question. Are members paying for the services they receive? Are service risks being disclosed? Are fees changing because of normal costs or crisis costs? Are committees acting within clear authority? Are legal constraints from Mauritius understood by resource members across the region?
Information rights also reduce the need for courts. If members can inspect the basis for a contested bylaw change, challenge a voting list before an election, see proxy rules before votes are cast, review aggregate legal spending before fees rise, and understand service impacts before rumours spread, fewer disputes need emergency judicial intervention. Courts remain available for legal questions. They should not be the first place members learn how the registry is governed.
The cost of information is modest compared with the cost of opacity. A registry can publish dashboards, audited financials, board minutes, risk registers, election-readiness reports, bylaw comparison notes, litigation categories and service metrics without compromising secrets. It can use independent auditors for member-register verification. It can require director and committee conflict disclosures. It can create a member ombuds or review panel for procedural complaints. These tools are less dramatic than receivership, but far more valuable. They keep accountability inside the membership relationship before crisis converts it into litigation.
The point is not to flood members with documents. It is to give them enough structured information to decide whether directors and management are acting as stewards of a narrow ledger or as claimants to an institution whose monopoly they control. A member should not need factional rumours to know whether services are stable, fees are justified, elections are prepared, litigation is bounded and resource-policy decisions are predictable.
For AFRINIC, rebuilding trust will require that kind of ordinary transparency. A board and budget are necessary. They are not enough. Members need to see, repeatedly and in usable form, how their monopoly ledger is being run.
A continuity firewall without institutional immunity
AFRINIC's defenders are right about one thing: continuity matters. A sudden failure of the African RIR would create real coordination problems. Members need records maintained, services kept online, RPKI and reverse DNS handled, resource requests processed, transfers evaluated and global registry relationships preserved. ICANN's intervention in a winding-up context, including its argument that numbering resources administered through AFRINIC should not be treated as corporate assets available for distribution, reflects a legitimate concern. A regional registry is not an ordinary company whose assets can be liquidated without attention to a public coordination function.
But continuity can be invoked too broadly. The fact that registry continuity matters does not mean every incumbent decision deserves deference. It does not mean members should accept weak elections, opaque legal spending, vague bylaw rights or discretionary enforcement. It does not mean courts should ignore resource holders' legitimate contractual complaints. It does not mean critics are wrong whenever their actions inconvenience the institution. Continuity is a function to protect, not a shield for unaccountable control.
The proper model is a continuity firewall. Essential ledger services should be insulated from corporate-control disputes, budget fights, board turnover and major litigation. If a board fails, the registry should still publish records and support security functions. If a member sues, ordinary services for unrelated members should not stop. If bank accounts are restrained, lawfully protected operating funds or emergency support arrangements should keep critical services running. If a board election is challenged, temporary authority should be limited to maintenance rather than sweeping policy or mandate changes. If liquidation is sought, courts should distinguish the corporate shell from the numbering function and the rights of resource holders.
AFRINIC's receivership provided an improvised version of this firewall, but it also exposed its limitations. The court could restrain takeover, preserve assets and require elections. Other RIRs and global bodies could express support. Staff could continue services. Yet the system still depended on a fragile mix of court orders, receiver discretion, election committees, external concern and public pressure. That is not a stable long-term design for a continent's number registry. Members should not have to wait for institutional collapse before continuity rules become visible.
Capture is the opposite risk. A registry can survive formally while losing member legitimacy. Capture may come from a commercial holder seeking favourable treatment, an incumbent staff culture resisting review, a government-linked bloc pursuing relocation or political control, a global coordination body trying to impose a preferred outcome, a faction using proxy votes, or directors using continuity rhetoric to avoid scrutiny. The continuity firewall must protect the ledger from all of these, not only from hostile litigation.
That requires design choices. Election rules should prevent sudden aggregation of voting power through unverified credentials. Bylaw changes should not be rushed while members are confused about their legal status. Litigation settlements that materially affect resource policy should require transparent board authority and, where appropriate, member consultation. Emergency support from other RIRs should be disclosed with conditions. ICANN or NRO intervention should be limited and explained so it protects continuity without becoming external occupation. Government or continental-body involvement should be visible and constrained by member rights.
The 2025 and 2026 episodes around AFRINIC demonstrate how narrow the path is. A board election restored formal governance, but critics questioned process and influence. AFRINIC said it was rebuilding budgets and strategy, but soon after warned of continuing litigation and procedural roadblocks. ICANN intervened in winding-up proceedings to protect registry continuity, but its earlier election interventions were themselves criticised as overreach by some observers. Bylaw reform aimed at legal clarity, but critics framed it as a reduction of member rights. Every repair risked being read as capture by one side or another.
That is why the answer cannot be trust us. It must be architecture. Continuity is protected when members can see what is being protected, who controls the protection, what powers are temporary, what rights remain with members, what services are ring-fenced, and how the institution returns to ordinary accountability. AFRINIC's crisis is the evidence that continuity without member control is too fragile.
The firewall has a second purpose: it makes accountability less dangerous. If members know that essential services are protected, they can challenge directors, scrutinise legal budgets and contest bylaws without fearing that the ledger will collapse. If every governance challenge threatens continuity, then accountability becomes hostage to operational fear.
What accountable membership would actually discipline
Membership accountability can sound vague unless translated into controls. In AFRINIC's case, it should discipline the member register, elections, budgets, litigation, enforcement, mandate boundaries and continuity planning. Each area corresponds to a failure mode already visible in the crisis.
The member register is first. AFRINIC needs a reconciled, court-defensible and member-visible distinction among resource members, associate members, registered corporate members if any, directors, authorised representatives, billing contacts, voting contacts and proxy holders. Members should be able to verify their own status and representatives before any election or meeting. Changes to voting authority should require direct confirmation through trusted channels. Powers of attorney should be narrowly defined, time-limited, validated and disclosed in aggregate. A registry that cannot reliably say who may vote cannot claim a stable member mandate.
Elections are second. Nominations committees should be appointed through a process whose conflicts are disclosed and whose authority is not improvised. Candidate eligibility criteria should be known in advance. Election operators should be independent, but their instructions should be public enough for members to assess. Online and in-person voting rules should match or be carefully reconciled. Incident procedures should specify when voting continues, pauses or is annulled. Any annulment should produce a member report describing scope, evidence categories, remedial steps and expected timeline. This is not over-engineering. AFRINIC's June 2025 annulment proved that election defects can prolong institutional paralysis.
Budgets are third. AFRINIC members should receive budgets that show core registry operations separately from community programmes, legal costs, exceptional recovery expenses, capital improvements, external advisers and reserves. Fee changes should be linked to those categories. Members should know whether they are paying for ordinary services, crisis defence, development activities or rebuilding. Board-approved budgets should be published without delay, and material deviations should be explained. A monopoly fee base deserves more than general assurances about financial sustainability.
Litigation is fourth. The registry should publish a litigation-risk register at a level that protects privilege but informs members. Major cases should be described by category, possible operational impact, expected cost band, board authority and next procedural milestone. Settlement principles should be documented. If legal action could affect transfers, allocations, membership rights, bank accounts or continuity, members should be told. If legal spending is delaying training or service improvements, that trade-off should be quantified. Legal silence may be prudent in a single lawsuit; systemic opacity is not prudent in a member-funded monopoly.
Enforcement is fifth. Resource reviews, contractual breach allegations, transfer denials, account closures, reverse DNS restrictions and RPKI-related decisions should have clear notice, evidence, appeal and proportionality rules. The registry must be able to enforce policies, but enforcement should not be a discretionary weapon. Members should know what conduct triggers review, what data is considered, how conflicts are handled, how long decisions take, and what independent forum exists if staff or board action threatens a resource position. The lesson from the Cloud Innovation dispute is that weak procedural confidence can turn enforcement into institutional warfare.
Mandate boundaries are sixth. AFRINIC should define which functions are core ledger functions and which are supplementary community activities. Core functions deserve continuity protection and priority funding. Supplementary activities need budget discipline and member approval. Policy implementation should be separated from political messaging. Regional-development claims should not justify hidden capital controls. IPv6 promotion should not obscure the continuing economic importance of IPv4 records. A member-funded registry may advocate, train and convene, but it should not blur those roles with its authority over recognised resource positions.
Continuity planning is seventh. Members should know what happens if the board fails again, if a chief executive cannot be appointed, if bank accounts are restrained, if a critical service provider fails, if a court order affects the registry, if a large member dispute threatens operations, or if emergency registry planning is discussed. The plan should identify ring-fenced services, emergency authorities, member notices, inter-RIR support arrangements, data escrow, financial reserves and limits on temporary powers. Continuity should be boring before it is needed.
These controls are not radical. They are the minimum institutional equipment for a compulsory ledger that administers scarce identifiers. AFRINIC's crisis is unusual in detail but not in principle. Wherever exit is unavailable and registry decisions affect value, member control must be made operational. The question is not whether a registry should be allowed to function. It must function. The question is whether its members can see and discipline the decisions that turn a technical ledger into an economic gatekeeper.
The rebuild must be judged by member control, not mood
AFRINIC's 2026 public posture included signs of recovery. Its new board had begun acting. Interim executive functions were being assigned. A budget and action plan were expected. A longer strategy for 2027-2030 was being discussed. Staff morale was reported to have improved. AFRINIC reappeared in regional internet forums after years of absence. Those are positive indicators. An institution that has been boardless and under receivership needs signs of ordinary administration.
But mood is not governance. Applause at a technical conference, optimism from staff, and the language of return are not substitutes for member-control repair. The correct test is whether the new board reduces the economic risks that made the crisis so damaging. Can members verify voting rights without confusion? Can they see a budget that separates legal recovery from registry operations? Can they evaluate the cost of ongoing disputes? Can they trust that transfer and enforcement decisions will follow predictable due process? Can they understand whether resource members have enforceable rights under the bylaws and Mauritian law? Can they inspect service-continuity planning? Can they discipline directors without threatening the ledger?
The answer remained uncertain because litigation and procedural conflict continued. AFRINIC warned in 2026 of a web of litigation and roadblocks. Cloud Innovation continued to challenge the institution, including through winding-up proceedings. ICANN intervened in the winding-up matter to explain AFRINIC's unique role and the nature of numbering resources. Disputes arose over statements about leasing and alleged court approval. Bylaw review itself became contested. These developments do not negate recovery, but they show that recovery is not a single event. A board can be seated while legitimacy remains under test.
The board's first accountability task is therefore not to win the narrative. It is to lower members' uncertainty. That means publishing enough about finance, services, legal exposure, bylaw proposals and election controls that ordinary operators no longer need factional channels to understand their own registry. It means acknowledging that members include critics, commercial holders, small operators, universities, carriers, governments and technical community participants whose interests diverge. It means refusing to treat disagreement as disloyalty while also refusing to let well-funded actors capture the institution through procedural exhaustion.
The second task is to narrow the registry's mandate. AFRINIC should be strongest where its role is clearest: maintaining accurate records, stable services, secure publication, fair resource administration and transparent policy implementation. It should be cautious where its role is politically or economically expansive: deciding the proper commercial model for IPv4 holders, using resource control to pursue broad development objectives, or treating institutional continuity as equivalent to board discretion. The narrower the mandate, the easier it is for members to hold the board accountable.
The third task is to make legal peace possible without making legal pressure profitable. A registry cannot allow one litigant to dictate policy by suing repeatedly. Nor can it pretend litigation is only sabotage when the underlying governance design has weaknesses. The board needs a litigation strategy that protects essential services, defends clear registry functions, corrects real procedural defects and seeks settlement where continued fighting imposes more cost on members than benefit. That judgment cannot be made in total secrecy when members fund the fight.
The fourth task is to repair the membership contract in law. If resource-member rights are vulnerable because of the gap between bylaws and Mauritian company law, the board should not hide behind legacy language. It should present options, explain consequences, obtain member input and adopt a structure that courts can enforce. Members may disagree over the right design, but ambiguity is worse. The registry cannot rebuild confidence on rights that collapse when tested.
The fifth task is to treat small members as the benchmark. Large holders and global bodies will always find ways to be heard. The small ISP, university network, data centre or regional service provider needs predictable fees, responsive support, clean records, voting clarity and a board whose legal battles do not consume the registry. If the system works for those members, it is likely accountable. If it works only for factions with lawyers, it is not.
AFRINIC's crisis is African in setting but general in lesson. Any institution that operates a monopoly ledger for economically important identifiers faces the same accountability problem. It may be incorporated as a nonprofit, speak in community language and rely on technical legitimacy. But when the ledger becomes necessary for business continuity and scarce-resource value, members are no longer ordinary subscribers. They are captive users of a settlement system.
That status changes the institutional bargain. The ledger must be narrow enough to be predictable, strong enough to resist capture, transparent enough to be trusted, and accountable enough that members can discipline the people who spend their fees and control their recognised positions. Courts and global coordination bodies can provide backstops, but they cannot replace the everyday legitimacy produced by clean membership governance. Receivership can preserve a registry for a time, but it cannot be the normal way a continent's operators obtain accountability. Derecognition policies and emergency registry planning may be necessary, but they are last-resort tools. The first defence is a membership system that works before collapse.
AFRINIC shows what happens when that system is too weak for the value it administers. Reported record corruption damaged confidence in ledger integrity. Scarcity made IPv4 recognition economically valuable. Enforcement against a large holder escalated into litigation that affected the institution itself. Bank-account freezes and board failure exposed the fragility of corporate continuity. Receivership kept the function alive but did not supply member consent. An election meant to restore legitimacy was suspended and annulled over voter-authority concerns. A later board gave AFRINIC a route back to normal governance, yet legal, bylaw and membership-rights disputes continued. Each episode pointed back to the same issue: members paid for and depended on a registry whose control mechanisms were not strong enough for its economic role.
The remedy is not to romanticise the community, demonise commercial holders, or place all trust in courts. It is to design the registry as a narrow ledger with accountable member control. Members must know who can vote. They must know what they are funding. They must know how legal strategy is authorised. They must know where enforcement discretion ends. They must know how scarce-resource policy is implemented. They must know what happens if the board fails. They must know whether their rights are recognised in the legal structure that governs the organisation. Those are economic safeguards, not procedural niceties.
AFRINIC may yet stabilise. A functioning board, a credible budget, clearer bylaws, stronger election controls and disciplined legal strategy would all reduce the premium now attached to its governance. But stability should not be measured only by whether services remain online or whether global bodies avoid emergency intervention. It should be measured by whether the small member who cannot change registries can see, understand and influence the institution that records its network identity.
That member is the reason accountability matters. It pays because it must. It relies because the market recognises only one regional ledger. It cannot diversify away from the registry in the way it can diversify upstreams, vendors or insurers. For that member, AFRINIC is not a club. It is the administrative layer beneath customer contracts, address value, routing security, reverse DNS, transfers and institutional recognition. A compulsory ledger can be legitimate only if the people compelled to use it can control it. AFRINIC's crisis has made that lesson impossible to treat as theory.