Why a registry election prices risk
An election at AFRINIC is not an internal club ritual. It is a pricing event for a scarce infrastructure ledger. The African Network Information Centre records who may use IP address space and autonomous system numbers across Africa and parts of the Indian Ocean. Its decisions affect routing confidence, transfer paths, member fees, RPKI and reverse-DNS services, allocation queues, and the credibility of a regional registry whose records are relied on by networks far beyond Mauritius, where the organisation is incorporated. The ballot therefore reaches beyond the meeting room into contracts, risk committees and network plans.
That is why board election legitimacy matters in a way that ordinary nonprofit governance often does not. A school association can survive a messy vote if members still accept the next budget. A registry that governs scarce number resources has a different problem. If members doubt that a board was produced by a clean electoral process, every later act by that board can acquire a risk premium. A budget may be treated as temporary. A CEO appointment may be attacked as tainted. A bylaw consultation may be read as consolidation. A transfer policy may be discounted by market participants because the authority behind it is still litigated. A registry database, whose value lies in being boring and final, starts to look provisional.
AFRINIC's recent history makes the point unusually concrete. The organisation operated without a board from 2022, was placed under receivership by the Supreme Court of Mauritius, attempted a board election in June 2025, saw that election suspended and then annulled after allegations around powers of attorney, and later announced a new board in September 2025. The Register reported that the new board gave AFRINIC a chance to convene directors for the first time in years, but also that critics questioned whether the election arrangements complied with the bylaws, that a criminal investigation into the June vote was under way, and that further court challenges were likely. In other words, the election did not simply end a governance vacuum; it became another object inside the dispute.
The economics begin with the asset under administration. IPv4 addresses are finite, still commercially necessary, and no longer obtainable at will from free pools. AFRINIC's own exhaustion material identifies IPv4 as scarce and records the region's move into Soft Landing Phase 2. The registry's fee schedule prices membership and allocation services; the market prices routable IPv4 very differently. That gap means registry discretion is valuable. Whoever controls the board influences the institution that sets budgets, appoints management, supervises policy implementation, responds to litigation, handles member standing, and frames the registry's position on transfers, regional use and commercialisation.
The usual governance vocabulary understates this. "Member confidence" is not only sentiment. It is a form of institutional collateral. A resource holder deciding whether to request an allocation, contest a compliance action, support a policy proposal, sign a power of attorney, lease address space, buy a business dependent on AFRINIC resources, or challenge a board resolution is pricing the chance that the registry's authority will be accepted. The same is true for ICANN, the NRO, courts, governments, creditors and counterparties. A legitimate election lowers transaction costs because parties can treat the board as a settled decision-maker. A contested election raises costs because every transaction carries an implicit governance caveat.
This article treats AFRINIC's board election legitimacy as a market and ledger risk. The issue is not whether one faction's preferred candidates should have won. It is whether the electoral system can produce a board whose authority is strong enough to protect the registry ledger, limited enough to avoid capture, and credible enough to govern a scarce resource without converting every policy dispute into a fight for corporate control.
The institution whose board became valuable
AFRINIC describes itself as a nonprofit, member-based organisation registered under Mauritius law and entrusted with distributing and managing internet number resources, including IPv4, IPv6 and ASNs. That description sounds administrative, and in a narrow technical sense it is. Number resources must be unique. WHOIS and RDAP records must be reliable. Reverse DNS, IRR and RPKI services must be maintained. Requests must be evaluated against policy. Members must pay fees so the registry can operate.
Yet the board of such an institution has become economically valuable because administration now sits on top of scarcity. A registry that once looked like a technical coordinator now stands between members and scarce inputs. It does not own the internet, and the official RIR position resists treating addresses as conventional private property. But the registry records the claims on which live networks, customer contracts and market transactions depend. A board that controls the organisation holding that record has influence over the rules of economic life around the resources.
Several functions create that influence. The board appoints or supervises executive management. It approves budgets and financial strategy. It determines how the organisation responds to lawsuits and settlement opportunities. It can shape bylaw reform and institutional posture. It can decide whether staff take a conciliatory or aggressive view of resource enforcement. It can frame public communications that either reduce or intensify disputes. Even where formal number-resource policy is made through a bottom-up policy development process, a board and management team matter because they execute policy, interpret procedure, allocate legal resources and decide what risks to fight.
For ordinary members this is not abstract. AFRINIC's fee page shows that membership categories and annual charges are tied to the resources held by members. Its exhaustion page shows that resource requests are evaluated through hostmaster processes and scarcity rules. Its policy manual treats number resources as public resources distributed according to documented need and registry rules. For a network operator, those official materials are not academic governance texts. They are part of the conditions under which a network can expand, transfer, maintain records, keep good standing and avoid interruption.
The board therefore carries a control premium. In corporate finance, control premium means the extra value attached to the ability to direct a company rather than merely hold a minority interest. AFRINIC is not a normal company with shareholders trading equity. But the analogy helps. Control of the board confers influence over an organisation whose registry function affects scarce resource value. That influence is not ownership of the resources, but it can affect liquidity, exit options, compliance risk and the perceived security of holding.
This is why a board election in AFRINIC can attract intensity out of proportion to the apparent administrative office. If the board were only approving stationery and meeting venues, election disputes would be embarrassing but not systemic. Instead, the board sits above institutional machinery that may validate or constrain commercial models. A member holding scarce IPv4 resources will care whether the board treats transfer mobility as a market function or a leakage from regional stewardship. A new entrant will care whether the board views remaining IPv4 pool management as conservative rationing or as a transition service. An incumbent critic will care whether bylaw reform strengthens members or narrows them into a less powerful category.
Heng Lu's public notes make a sharper version of this point. In his view, elections validate power more often than they create it, because the decisive control lies in procedures, capital, networks and institutional leverage. That is an interested perspective from a participant in AFRINIC disputes, but the underlying institutional lesson is useful: when control over process can affect control over scarce assets, the process itself becomes an economic asset. Legitimacy is not decoration around power. It is the mechanism that makes others accept the board's later use of power.
Scarcity turned representation into a market question
The RIR model depends on representation by participation. Members, operators and other stakeholders are expected to use policy meetings, mailing lists, committees, elections and consultations to shape the rules. This works tolerably well when the affected community is small, technically literate and willing to spend time on institutional detail. It becomes fragile when the number of economically affected parties is larger than the number of people who actually participate.
AFRINIC's service region is broad and heterogeneous. It covers dozens of countries across Africa and the Indian Ocean. Public descriptions of the organisation divide the region into sub-regions for board representation and include regional and non-regional board seats. That design is sensible as a map. It is less obviously sufficient as a mandate. A sub-regional seat does not guarantee that telecom operators, ISPs, datacentres, government networks, universities, banks, cloud platforms and address holders in that sub-region have comparable interests or similar incentives to participate.
The representational gap matters because scarcity makes abstention costly. When IPv4 was easier to obtain, a member could ignore governance because registry politics rarely changed the value of an existing operational input. Once IPv4 became scarce and transferable or leasable in practice, governance choices began to affect balance sheets. A policy that restricts out-of-region mobility can reduce liquidity. A compliance interpretation can threaten a revenue stream. A bylaw change can alter voting rights. A board election can determine whether the registry treats market participants as legitimate operators, opportunists, or threats.
The public record contains repeated signs that many members do not behave like active shareholders in a high-stakes institution. ISPA's March 2025 warning, reported by The Register, told South African members to be vigilant about AFRINIC credentials because entities obtaining multiple members' credentials could manipulate voting and alter board composition or policy outcomes. AFRINIC itself had warned members about solicitations to access credentials by obscure or fictitious organisations. NRS, from the other side of the dispute landscape, later used public messaging to urge members to protect their vote and report cases where votes may have been recorded in their name.
These signals point to a simple institutional problem: passive members create room for organised intermediaries. Some intermediaries may help members understand rights they otherwise ignore. Others may aggregate votes in ways that weaken genuine consent. In a lightly attended governance system, a disciplined bloc can acquire outsized influence. That is not unique to AFRINIC; it is a chronic weakness in associations where members are operationally busy and governance specialists are persistent. But AFRINIC's scarcity context raises the stakes because control over a sleepy electorate can become control over valuable policy discretion.
Representation also becomes difficult because the phrase "the African internet community" can hide conflicting economic interests. An access provider seeking cheap IPv4 capacity is not identical to a holder seeking transfer value. A government ministry concerned with national digital plans is not identical to a private datacentre operator worried about customer continuity. A standards participant may prefer conservative stewardship; a leasing company may prefer mobility. A sub-regional seat does not dissolve these conflicts. Nor does continental language. It may even obscure them by implying that one slate or one institution can speak for a region whose networks face different risks.
The September 2025 board election highlighted that tension. The Register reported that seven of eight elected directors had Smart Africa's endorsement. Smart Africa is a government-linked continental digital-development organisation with many member states. It has called for a coordinated response against institutional capture and disruption of critical internet functions. That concern is legitimate on its own terms. But The Register also reported discomfort among some in Africa's internet community about a Smart Africa-backed majority, including concerns that a few countries might dominate and that some might be interested in AFRINIC's institutional home.
The point is not that Smart Africa's endorsement made the board illegitimate. Endorsements are normal political facts. The point is that representational legitimacy is not produced merely by invoking the continent, the community or stability. It must be shown through voter eligibility, informed consent, clean authorisation, equal rules, transparent counting, auditable registers and credible dispute handling. A board can represent a service region only if the path by which it was elected is stronger than the factions competing to claim the region's voice.
The receiver's narrow mandate and the danger of legislative repair
Receivership was supposed to preserve the institution while a board was restored. The NRO's September 2023 statement, based on the Mauritius court process, described the receiver's role as maintaining the status quo of AFRINIC's assets, preserving the value of the business, overseeing elections in accordance with AFRINIC's constitution, facilitating a proper board and appointing a chief executive. It also noted restrictions on relocation, takeover, merger, restructuring or management control. That factual exhibit matters because it defines the repair mechanism as preservative.
A preservative mandate is necessarily narrow. It keeps the lights on, protects the registry ledger, prevents opportunistic corporate changes and creates a lawful path back to member governance. It does not itself answer the underlying policy questions that made control valuable. A receiver may need to organise voting, verify eligibility and keep services running. But the more a receiver appears to reshape institutional power, reinterpret member rights or enable structural policy moves while legitimacy remains contested, the more the receiver becomes part of the dispute rather than the bridge out of it.
That distinction is central to board election legitimacy. AFRINIC was not merely short of directors; it was short of accepted authority. A receiver-run election had to do more than produce names for board seats. It had to demonstrate that a court-supervised repair process could create a board whose legitimacy would survive the next conflict. That requires a high evidentiary standard for election mechanics, because the receiver is operating in a context where every party already expects procedural advantage to matter.
The April 2025 election plan reflected this sensitivity. The Register reported that receiver Gowtamsingh Dabee appointed Simon Davenport KC and other UK barristers to a nomination committee, citing concerns about potential interference and the need for a free and fair election mechanism. Civica Election Services was chosen to conduct the poll, and an election committee included AFRINIC executives, a Civica representative and an accountant from the receiver's office. Those design choices suggest the receiver understood that ordinary process would not be enough. Legitimacy required visible independence.
Even so, independence on paper does not solve authorisation risk. A nominations committee can vet candidates, but it does not by itself verify that every vote reflects the will of the member entitled to cast it. A professional voting vendor can run a poll, but it cannot rescue inconsistent proxy rules. A court can decline to halt an election, but that does not mean the resulting board will command market confidence if members later dispute whether votes were cast with consent. Electoral outsourcing can improve mechanics; it cannot manufacture member consent after the fact.
The receiver's position is therefore delicate. Moving too slowly keeps AFRINIC in paralysis. Moving too quickly can produce a board vulnerable to attack. Relying too much on legal form can miss operational realities, such as how resource members manage credentials, how powers of attorney are solicited, and how staff records are maintained. Relying too much on pragmatic fixes can invite claims of ultra vires action or bylaw evasion. In a normal association these risks might be absorbed. In AFRINIC they compound because the election is the gateway to every subsequent act of institutional recovery.
Heng Lu's February 2026 note used forceful language to argue that a receiver is preservative, not legislative, and should not be used to redefine the economics of member-held resources while board legitimacy is unresolved. One need not accept every conclusion in that note to see the governance problem. If a board elected through disputed mechanics quickly ratifies economically consequential policy, sceptical members will not view the election as a cure. They will view it as the enabling event for control. The board's authority then becomes inseparable from the contested process that produced it.
Voting design and the proxy problem
The June 2025 election failed at the exact point where representation becomes evidence: voter authorisation. The Register's June 26 report stated that online voting rules allowed one member to hold proxies for five others, while rules for in-person voting allowed one person to vote for any number of members if they held a power of attorney for each. That asymmetry was not a minor technicality. It created an incentive to shift influence into the channel with weaker aggregation limits.
Proxy voting is not inherently illegitimate. Many members cannot attend meetings. Multinational operators may centralise governance tasks. Smaller networks may rely on advisers. In a region-spanning registry, some form of delegation is practical. The problem is that proxy systems turn identity verification into the heart of the election. The registry must know who the member is, who is authorised to act, whether the authorisation is current, whether it was voluntary, whether it has been revoked, and whether the same member has already voted through another channel.
AFRINIC's annulled election exposed weaknesses in that chain. The nomination committee suspended the vote minutes before the in-person period ended because of questions around the validity of powers of attorney or member-given voting powers. ISPA then alleged that authorised representatives of members had arrived to vote only to discover that someone else had already submitted a vote using a power of attorney the member had not provided. ISPA also alleged that election officials could not produce a relevant power-of-attorney document when asked. AFStar reportedly alleged that two powers of attorney had been proven fraudulent. ICANN's subsequent letter referred to allegations of fraudulently obtained powers of attorney and asked the receiver to answer a series of questions.
The strongest version of the problem appeared in The Register's July 11 follow-up. ISPA told the publication that one party claimed powers of attorney representing nearly half of all AFRINIC resource holders. ISPA said evidence emerged that at least some were fraudulent. An anonymous AFRINIC member told The Register that someone had tried to vote on his behalf using a document he said he had never signed. ICANN also claimed the receiver had discovered one forged power of attorney. AFRINIC, the receiver and the nomination committee did not provide public detail sufficient to close the confidence gap.
In institutional-economics terms, the election suffered an authorisation shock. The nominal vote count no longer signalled consent. If a material share of votes could be cast through disputed or unverifiable authorisations, the election did not merely produce a contested outcome; it damaged the credibility of the membership register. The register is the same broader institutional object on which the registry's legitimacy rests: a set of records telling the world who has which rights, who can speak for which organisation, and which changes are valid.
The link between voting records and number-resource records is not direct, but it is psychologically and institutionally powerful. If members doubt that AFRINIC can verify who may cast a vote, they will ask whether it can verify who may request a transfer, update a resource record, sign an RSA, lodge a complaint, or authorise a policy representative. That is why election fraud allegations are more dangerous in a registry than in many associations. They raise doubts about the competence of the organisation's identity and authority controls.
The proxy design also created a strategic narrative for every faction. Critics of the June process could argue that organised vote capture was under way. Defenders of annulment could argue that the receiver protected legitimacy by stopping a corrupted vote. Critics of annulment could argue, as Cloud Innovation did publicly, that a narrow dispute was being used to void the collective voice of members and perpetuate instability. Each narrative contains a plausible institutional concern. That is precisely why the underlying evidence had to be published with unusual care. Silence allowed every side to price the facts in its own favour.
Annulment, silence and the value of explanation
The receiver annulled the June 2025 election after the suspension. The stated reason, as reported by The Register, was feedback and concerns from stakeholders about potential irregularities related to voter documentation, with a goal of protecting transparency, fairness and unquestionable legitimacy. That may have been the right decision. A board elected through fraudulent or unverifiable authorisations would have been worse than a delay. But annulment without a clear public account carried its own cost.
In high-stakes governance, explanation is not public relations. It is institutional settlement. An annulled election creates losers, and losers need to know whether the decision was grounded in facts, law and proportionate reasoning. Members need to know whether their own vote was counted, duplicated, replaced or rejected. Candidates need to know whether the process failed because of fraud, poor rules, administrative confusion, or tactical challenge. Courts need a record. Market participants need to know whether the next election will fix the defect or merely replay it.
ICANN understood this, even if its own interventions were contested. Its July 2025 letter criticised the receiver for not explaining the annulment or responding adequately to transparency concerns. It demanded a transparent reporting of the investigation and findings. It also referred to the emergency framework under which ICANN might review AFRINIC compliance and eventually seek emergency registry arrangements if necessary. Official global oversight should not be treated as the source of AFRINIC's legitimacy, but its alarm is useful evidence of how serious the annulment looked from outside.
The absence of a detailed public account widened the bid-ask spread on legitimacy. Buyers and sellers of institutional confidence could not agree on the price because the facts were not cleared. One side could treat the annulment as proof of electoral corruption. Another could treat it as proof that any contested ballot could be weaponised to prevent a board. A third could treat it as evidence that the receiver lacked a firm grip on election administration. With no authoritative post-mortem, the election became not an event with lessons but an unresolved claim.
This matters because later elections borrow credibility from earlier repairs. If the June defects were specific, documented and fixed, the September election could be viewed as a corrected process. If the June defects were never fully explained, September becomes harder to assess. Were the same voter lists used? Were powers of attorney revalidated? Were proxy limits harmonised? Were member credentials reset? Were staff able to audit authority records? Did disputed members receive notice and a channel to contest? Without public answers, even a successful vote remains exposed to challenge.
The Register's September 2025 report captured the result. AFRINIC had elected eight directors, allowing the organisation to convene a board, hire leadership, seek to unfreeze bank accounts and resume work. But critics claimed the election arrangements might not be allowed under bylaws. Stakeholders were expected to ask Mauritian courts to consider whether the election was properly run. Government and criminal investigations hovered around the wider process. The registry had crossed the formal threshold of electing directors without crossing the confidence threshold of putting legitimacy beyond serious dispute.
The lesson is not that every election controversy requires a public dump of sensitive member documents. Privacy and legal process matter. But a registry should be able to publish enough aggregate and procedural evidence to establish confidence: the number of challenged authorisations, the categories of defect, the verification method, the rule changes made before the next vote, the appeal path, and the treatment of affected votes. In a ledger institution, unexplained correction is only half a correction. The other half is a record that others can rely on.
The September board and the control premium
The September 2025 board election changed AFRINIC's formal posture. A board existed. The organisation could begin taking actions that had been blocked by the absence of directors. The Register's February 2026 coverage reported improved morale, interim management appointments, budget and action planning, and a strategic planning horizon for 2027 to 2030. AFRINIC's staff message at APRICOT 2026 presented the registry as emerging from the quagmire. For operators who need ordinary registry services, that matters. A flawed board may still be operationally better than no board.
Yet formal capacity is not the same as uncontested authority. The newly elected board arrived with immediate economic questions around its mandate. Would it preserve the registry as a narrow ledger? Would it endorse stronger regional lock-in? Would it resist commercial leasing? Would it settle or intensify litigation with Cloud Innovation and related entities? Would it support bylaw reforms that clarify or reduce resource-member rights? Would it align with Smart Africa's continental-stability agenda, the existing RIR community's derecognition framework, or member demands for more direct asset protection?
Each question is a control-premium question. The value of board control lies not only in chairing meetings but in choosing the institutional path through unresolved trade-offs. For example, The Register's March 2026 report described AFRINIC accusing Cloud Innovation, Larus and associated advocacy campaigns of creating litigation and procedural roadblocks. The same report quoted Lu Heng arguing that the deeper issue was a registry model concentrating high-consequence power over economically critical resources without commensurate liability. AFRINIC's side framed litigation as paralysis. Heng's side framed registry discretion as structural risk. The board's posture determines which narrative becomes policy.
The regional transfer issue illustrates why control is priced. AFRINIC has adopted or considered policies that restrict the mobility of resources issued in its region. Supporters see this as protecting African number resources from extraction and preserving regional development. Critics argue that it traps holders, destroys liquidity, depresses asset value and increases registry dependency. A board that treats regional lock-in as stewardship will make different choices from one that treats mobility as a governance discipline. That difference can be worth millions across large holdings and can alter the perceived security of smaller operators.
Legitimacy therefore acts as the licence for structural change. A board elected by a widely trusted process can make hard policy choices and ask losers to accept them because the decision path was fair. A board elected through a contested process may make the same choices and face claims that it is laundering control through institutional form. That is why a powerful electoral mandate can be double-edged. If a slate appears to win overwhelmingly in a fractured, recently annulled, litigation-heavy environment, supporters may see consensus. Sceptics may see capture, especially if member reports later suggest votes were recorded without consent or if the authorisation system remains opaque.
The control premium also explains the behaviour of external actors. ICANN's concern is not merely whether AFRINIC has directors, but whether those directors can maintain global coordination of the numbering system. The NRO's concern is continuity among RIRs and the ability to avoid a precedent in which one registry's collapse destabilises the whole model. Smart Africa's concern, as reported, is preventing institutional capture and disruption of critical African internet functions. NRS's concern is registry chokepoint power over members' money, records and votes. Each actor sees board control as a lever over a different risk.
In a healthier institution those risks would be mediated through predictable procedure. AFRINIC's problem is that procedure itself is contested. The board's first years will therefore be judged less by slogans about renewal than by whether it can reduce the control premium attached to its own seats. The best sign of recovery would be a board that makes itself less economically valuable: narrower discretion, published records, clearer member rights, defined proxy rules, separation between ledger administration and policy punishment, and enough restraint to show that no faction needs control of AFRINIC to protect itself from AFRINIC.
Ledger risk for members, buyers and downstream networks
Board election legitimacy is a ledger risk because the registry's records are valuable only if parties believe the institution behind them is stable, constrained and procedurally competent. A database can be technically online while institutionally impaired. WHOIS can return a holder name even if courts, members or counterparties doubt whether the governance system can process changes without dispute. RPKI and IRR services can function while resource holders worry that policy decisions will later invalidate commercial expectations.
For existing members, the immediate risk is continuity. If a board's authority is challenged, members may face uncertainty about invoices, standing, transfers, resource requests, name changes, or compliance processes. A member that wants to restructure corporate holdings may hesitate if registry approval could be caught in litigation. A member needing additional IPv4 capacity may confront delays because AFRINIC's remaining pool, staff processes or policy implementation are constrained by legal uncertainty. A holder considering a transfer must price the chance that the registry will refuse, delay or condition the transaction based on disputed policy.
For buyers and lessors, the risk is title-like confidence without formal title. IPv4 transactions often rely on a chain of contractual assurances and registry updates. The registry does not create a conventional property deed, but it records the authoritative operational state. If that state can be interrupted by board litigation, bylaw disputes, receiver questions, or contested policy, the market discounts the resource. Blocks associated with a stable registry command more confidence than blocks whose administrative path may be challenged. Even when the address itself routes normally, the governance discount can appear in price, contract terms, indemnities and willingness to transact.
For downstream networks, the risk is more practical. Customers rarely care which RIR recorded the resource. They care whether routing remains stable, abuse contacts work, geolocation is manageable, reverse DNS can be maintained, RPKI does not break, and a service provider can keep offering addresses. If registry-layer conflict causes a holder to lose recognition, face transfer denial, or become trapped in litigation, downstream customers absorb operational costs. Renumbering, firewall updates, reputation resets, allow-list changes and contract disputes are expensive. The registry's election legitimacy therefore travels down the stack into customer continuity.
KrebsOnSecurity's 2019 reporting on alleged historical address-record manipulation at AFRINIC shows why ledger trust is not an abstraction. The report described allegations that dormant or defunct African address blocks had been commandeered and sold, with an estimated market value above USD 50 million for documented addresses. AFRINIC's then chief executive said an investigation was under way. Whatever the full legal and factual resolution of those allegations, the market lesson is straightforward: weak record controls around scarce IPv4 can turn administrative uncertainty into high-value economic harm. Election records and resource records are not the same system, but both depend on verified authority.
The election authorisation problem is a cousin of that ledger problem. In both cases, the institution must know who is entitled to act. For resource records, it must know who holds or controls a block. For elections, it must know who can vote for a member. For transfers, it must know who can authorise movement. For policy, it must know who is genuinely participating. A registry with unreliable authority verification invites both fraud and capture. A registry that overreacts with discretionary control invites litigation and value destruction. The balance is narrow.
The best registry ledger is not one that wins every policy fight. It is one that makes contested facts manageable. It distinguishes record correction from punishment, fraud from commercial disagreement, member consent from proxy aggregation, and board authority from factional endorsement. AFRINIC's board election legitimacy matters because a board produced by a trusted process can restore those distinctions. A board produced by a distrusted process may blur them further, making every ledger act look like an act of power.
Courts, ICANN and the local wrapper problem
AFRINIC is a Mauritian company performing a regional and global coordination function. This dual character explains much of the crisis. Courts in Mauritius have ordinary authority over the corporate entity. ICANN and the NRO have systemic concerns about the registry function. Members have contractual and governance claims. Governments have public-interest concerns. Network operators have operational reliance. When the board is missing or contested, all these authorities press against the same institutional shell.
The receivership order showed that local law can protect the system. The court-appointed receiver was a legal backstop when internal governance could not produce a board. IGP's October 2023 analysis framed receivership as evidence that private internet governance can self-correct through rule of law and ordinary legal safeguards. That is an important point. The alternative to court-supervised repair might have been unilateral intervention by other RIRs, political takeover, or indefinite paralysis.
But local legal repair also has limits. A Mauritius court can appoint a receiver, interpret company law, order communiques and supervise winding-up applications. It cannot by itself produce trust among every AFRINIC member or resolve the political economy of IPv4 scarcity. It may need evidence from ICANN to understand why number resources are not simply corporate assets available for distribution in liquidation. It may need to hear from litigants whose private rights are affected. It may need to distinguish the registry corporation from the registry function.
The Register's May 2026 report showed this local-wrapper problem in a new form. ICANN was allowed to become a party to Cloud Innovation's application to wind up AFRINIC. ICANN's stated purpose was to explain AFRINIC's unique role and to make clear that numbering resources allocated through AFRINIC are not AFRINIC assets available for distribution in a winding-up. That intervention is not a conclusion source for this article; it is a factual exhibit of the legal problem. A local court was being asked to consider the fate of a company whose records affect global internet coordination.
This dynamic raises a legitimacy risk of its own. If external institutions intervene too aggressively, members may view AFRINIC as no longer governed from within its service region. If they intervene too little, local corporate litigation may endanger the registry function. If courts prioritise corporate form, they may miss systemic reliance. If global bodies prioritise systemic continuity, they may appear to override member rights. A clean board election would not eliminate this tension, but it would reduce the need for external actors to fill a legitimacy vacuum.
ICANN's 2025 election interventions reveal the calibration problem. Before the June vote, ICANN raised concerns about nomination committee conflicts and the erroneous listing of Cloud Innovation in corporate records. The Supreme Court declined to reconstitute the nomination committee and noted that ICANN lacked standing, while a communique clarified the corporate-record error. After the election was suspended and annulled, ICANN demanded answers and warned of possible compliance review. One intervention was rebuked; another looked prescient after proxy allegations. The pattern shows the difficulty of external oversight when internal election legitimacy is weak.
For market participants, the practical result is uncertainty over which authority will be decisive next. The board, the receiver, the court, ICANN, the NRO, a government ministry, a criminal investigation, or a winding-up application can each affect the operating environment. That multiplicity is itself a risk premium. A legitimate board cannot make local law irrelevant, but it can provide a central point of institutional accountability. Without it, the registry becomes a forum in which every external authority competes to define stability.
What a legitimate AFRINIC election must prove
AFRINIC does not need a perfect election. No membership organisation has one. It needs an election good enough that reasonable losers accept the board as authorised while preserving their right to challenge specific policies later. That standard is higher than merely counting ballots and announcing directors, because AFRINIC's recent history has shown exactly where the weak points are.
The member register must be auditable. The registry should be able to show how many resource members are eligible, what categories of member can vote, how corporate representatives are verified, and how disputes over authority are handled. The detailed personal or corporate documents need not all be public, but the process and aggregate results should be. If members suspect that the voting roll is stale, altered or unclear, no result will settle the question of representation.
Proxy and power-of-attorney rules must be uniform, limited and verifiable. The June 2025 asymmetry between online proxy limits and in-person power-of-attorney aggregation was a design failure even before any document was alleged to be fraudulent. Delegation should be possible, but not in a form that lets one actor appear with authority for a large share of the electorate without extraordinary verification. Every authorisation should be time-bound, specific to the election, revocable, confirmed through an independent channel, and visible to the member before the vote closes.
The system should provide a member-facing receipt. A resource member should be able to confirm whether a vote has been recorded in its name, through which channel, by which authorised representative, and at what time. If a member has not voted, it should be able to see that no vote has been lodged. This is not difficult in principle; banks and corporate registrars manage comparable authorisation trails. In a registry whose job is authoritative records, electoral records should be treated with similar seriousness.
The election authority must publish a post-election assurance report. It should describe challenges, rejected authorisations, duplicate attempts, late revocations, appeal outcomes and rule deviations. Sensitive documents can remain sealed, but the confidence story must be public. The June annulment failed partly because the public did not receive enough to understand the size and nature of the defect. A future election that says "trust us" will not repair a system whose problem is trust.
The board should observe a restraint period for structural economic policy after a contested election. This does not mean paralysis. Budgets, staffing, security, audits and routine registry services must proceed. But major changes affecting transfer mobility, resource-holder rights, bylaw identity, or enforcement posture should be supported by clearly documented authority and enough consultation to show that the new board is not using electoral form to entrench a factional victory. The more contested the election, the more important this restraint becomes.
Courts and external bodies should be used to support evidence, not to replace member legitimacy. Judicial supervision can validate procedures, protect records and resolve disputes. ICANN and the NRO can explain systemic consequences. They cannot make members feel represented if the election design itself is weak. AFRINIC's legitimacy must be rebuilt primarily through verifiable member consent, not through emergency endorsement from outside.
Finally, the board should reduce its own control premium. It can do this by separating registry ledger administration from policy advocacy, publishing clearer enforcement criteria, improving appeal mechanisms, and making member rights legible. A board that narrows discretionary power will face less incentive for factions to capture it. If control is less valuable, elections become less existential. That is the healthiest outcome for a registry.
Uncertainty and watchpoints
Several facts remain uncertain or contested. Public reporting describes allegations of fraudulent powers of attorney, member votes recorded without consent, missing documents, nomination committee conflict concerns, disputed bylaw compliance, government investigation, criminal investigation and continuing court cases. Not every allegation has been tested in a public judgment. The receiver may have information that was not fully disclosed. ICANN may have received correspondence not posted publicly. AFRINIC may have internal records that support decisions more strongly than the public record shows. Cloud Innovation, Larus, NRS, ISPA, Smart Africa and other participants all speak from positions affected by the outcome.
That uncertainty should not lead to false neutrality about the institutional problem. The problem is visible even if some allegations fail: AFRINIC's board election legitimacy remains economically consequential because control of the board affects a scarce-resource ledger, and the election machinery has already failed once in public. The unresolved question is whether the institution can now generate enough evidence, restraint and service delivery to narrow the legitimacy discount.
Start with the treatment of the September 2025 board in Mauritian courts. If courts uphold the election process and reject major challenges, the board gains formal stability. If courts find bylaw defects or procedural irregularities, every subsequent board action may require revalidation. Even if the board survives, the reasoning matters. A narrow procedural ruling may not satisfy members unless it addresses authorisation integrity and voting design.
The criminal or official investigations connected to the June 2025 vote are the next test. A finding that fraudulent documents were material would support the decision to annul but would also demand stronger controls before any future vote. A finding that irregularities were minor would raise questions about whether annulment was proportionate. No public finding, or a finding without operational detail, would leave the confidence gap open.
Bylaw reform will show where power is being moved. AFRINIC's future governance depends on whether resource members are given clearer rights or moved into a more limited participatory category under Mauritius company law. The Register's May 2026 reporting on ISPA's bylaw review highlights the tension between registered members, resource members and community-resolution mechanisms. Any reform that appears to reduce resource-member authority without overwhelming procedural legitimacy will be read as consolidation.
Transfer and regional-use policy remain the market-facing signal. Policies that mark resources as regional, restrict inter-RIR mobility, or constrain commercial leasing may be defended as stewardship, but they are also economically distributive. If such policies are adopted or enforced by a board whose election legitimacy remains contested, they will intensify litigation and market discounting. If they are debated through a cleaner, transparent process with serious economic evidence, losers may still object, but the registry's authority will be stronger.
Operational recovery is the most concrete legitimacy measure. Budgets, audited accounts, staff morale, service-level performance, allocation processing, RPKI and IRR stability, and member support are practical legitimacy measures. A board that restores services earns trust in a way no communique can. Conversely, a board that spends its early mandate mainly in litigation, public denunciation and bylaw fights will reinforce the view that control of the registry remains the prize.
The winding-up litigation and ICANN's participation will test the local/global balance. If courts treat AFRINIC primarily as an ordinary company, the registry function could face destabilising remedies. If courts understand the global public function but ignore member grievances, critics will argue that systemic language is being used to shield the incumbent institution. The stable path is one that protects the ledger while making governance accountability real.
The final watchpoint is member behaviour. If resource members remain passive, the next election will again be vulnerable to proxy aggregation, slate politics and procedural capture. If members become active only through factions that collect authorisations at scale, the same problem returns under a different banner. AFRINIC's legitimacy will improve when ordinary operators treat governance as part of infrastructure risk management: verifying contacts, protecting credentials, reading policy proposals, demanding receipts, and voting directly where possible.
AFRINIC's board election legitimacy is therefore not a settled event from September 2025. It is an ongoing market signal. The signal strengthens when the board acts with restraint, records are audited, courts clarify rather than improvise, members can verify their own authority, and policy decisions are separated from factional control. It weakens when unexplained process defects are followed by economically consequential decisions. The scarce asset is IPv4; the scarcer institutional asset is trust that the registry ledger is governed by consent rather than capture.