Summary
- AFRINIC issued election rules, court updates, committee announcements, voter instructions, registers, and result notices during its board vacuum. Those disclosures were necessary, but publication did not itself create a quorate board or cure defects in authority, eligibility, remedies, and execution.
- The failed June 2025 election demonstrated the limit most clearly: a heavily documented timetable still ended in suspension and annulment because the public record did not contain a sufficiently decisive method for isolating disputed authority documents, measuring their effect, and preserving valid votes.
- The board returned after a different combination converged in September 2025: a court-backed deadline, a replacement election design, direct voter designation, public register stages, external electoral supervision, a completed ballot, declared winners, and an operational handover. Transparency made those steps inspectable; it did not substitute for them.
Publication is evidence, not a corporate organ
An institution in crisis often promises transparency before it can promise resolution. The choice is understandable. A notice can be issued while litigation continues. A timetable can be displayed before every dependency is secure. A committee can be named before it has confronted its hardest evidentiary dispute. Publication is comparatively quick, visible, and reversible. Reconstituting a board is none of those things.
AFRINIC's experience from 2022 through 2025 makes the distinction unusually clear. The organisation published election material, court documents, receiver communications, voter requirements, candidate information, registers, deadlines, and eventually results. Interested observers could watch many parts of the institutional struggle in public. Yet AFRINIC still spent nearly three years without a functioning board. The quantity of disclosure did not shorten the vacancy by itself.
That is not an argument for secrecy. Without publication, members could not know the stated rules, test their eligibility, examine candidate choices, challenge deadlines, or compare institutional claims with court orders. The public record is one reason the crisis can be analysed at all. Transparency is a precondition for accountable recovery.
It is not the act that completes recovery. Under AFRINIC's constitution, directors had to be lawfully selected and take office. The eligible electorate had to be identified. Candidate and voter decisions had to be made by competent bodies. The ballot had to produce an outcome under rules that could withstand timely challenge. Corporate authority then had to move from temporary custodianship to the elected organ. A webpage could report each event, but it could not perform any of them merely by existing.
The central governance error is to treat disclosure as constitutive. A transparent announcement can describe authority; it does not create authority unless the governing instrument gives the announcement that effect. It can state a deadline; it does not supply a remedy when the deadline fails. It can list voters; it does not prove the chain from legal member to authorised human. It can announce winners; it does not settle every defect in the election that produced them.
AFRINIC therefore offers a general rule for critical infrastructure institutions: publication must be joined to an executable rights chain. The reader should be able to identify not only what was announced, but who had power to decide, what evidence controlled, how an error could be corrected before becoming irreversible, what happened if a responsible actor missed a date, and which event legally changed the institution's state.
The vacuum began as an authority failure
AFRINIC's board problem was not initially a communications problem. The organisation's own later history places the loss of a quorate board in June 2022. The Supreme Court judgment reported as Benjamin Adzenyamebeye Eshun v African Network Information Centre (AfriNIC) Ltd, 2023 SCJ 63, addressed the consequences of board composition and member rights in the unstable period that followed. By September 2023, the later appellate record treated AFRINIC as having no lawfully appointed director in office and no quorate board capable of authorising acts in the ordinary way.
Those facts mattered because a company cannot talk itself into a board. Directors are not produced by the continued presence of staff, publication of minutes, maintenance of services, or invocation of a regional community. The legal and constitutional steps for appointment remain decisive. If those steps cannot be performed through ordinary organs, another competent source of temporary authority is needed to preserve the company and open a lawful route back.
The Supreme Court's 12 September 2023 receivership order supplied such a route. Its preservation logic was commonly described as holding the ring, maintaining the status quo, protecting the value of the business, and facilitating the restoration of corporate governance. The Court of Civil Appeal's 2024 judgment, 2024 SCJ 473, later dealt directly with the authority to act for AFRINIC, restored the receivership order, and stressed completion of the board-election path.
These judicial acts were public and therefore transparent. Their operative importance did not arise from publication, however. It arose from lawful coercive authority. The receiver could manage and preserve the company because of the court order and applicable Mauritian law, not because a notice described the appointment. The same distinction applied to the effort to elect directors. A public election date could coordinate action, but the power to convene and supervise in the vacuum had to be anchored in the order, the constitution, and the receiver's lawful mandate.
This is why the phrase “more transparency” can be too small for the problem. If the defect is uncertainty over who may bind the company, the answer must identify a competent decision-maker. If the defect is an absent board, the answer must create a valid selection path. If the defect is a disputed election record, the answer must preserve evidence and provide a remedy. Disclosure makes those solutions reviewable. It cannot replace them.
AFRINIC had transparency of events before transparency of control
Transparency is often treated as one quality, but institutional recovery requires several. Event transparency tells the public that something occurred: a case was filed, a receiver was appointed, a committee was formed, or a vote was scheduled. Rule transparency states the criteria and procedure. Evidence transparency shows enough of the factual basis to test whether the rule was applied. Authority transparency identifies who could make the decision and under which instrument. Remedy transparency explains how a wrong decision can be corrected in time. Outcome transparency reconciles the inputs, exceptions, and final state.
AFRINIC produced substantial event and rule transparency. It issued communiques and maintained election pages. It published timetables, committee appointments, eligibility language, representation requirements, candidate slates, voter-register information, and results. The court record also exposed important changes in formal authority.
The harder layers were less complete. In the June 2025 election, the public could read detailed documentary requirements for electronic voters, proxies, and in-person representatives. But when concerns arose about authority documents, the public record did not contain a mature, pre-written law of materiality. It did not show a decision tree for separating a questionable authorisation from an otherwise valid member entitlement, deciding whether the concern affected one contest or the entire ballot, preserving undisputed votes, and issuing reasons proportionate to the evidence available.
After annulment, the receiver's 15 July notice gave a more specific attributed account. It identified suspicions particularly concerning powers of attorney, stated that police were investigating, acknowledged the absence of final conclusions, and explained why the receiver could not yet report the extent of the irregularities. That disclosure was more informative than a generic appeal to integrity. It also revealed why transparency alone had not solved the problem: the decisive facts were incomplete, the effect was not bounded, and the first election had already been extinguished.
The public had been told what happened. It still could not replay why the remedy had to reach every valid voter. The gap did not prove that annulment was improper; confidential evidence or urgent preservation concerns may have been weighty. It showed that event disclosure after an irreversible decision is weaker than evidence and remedy rules written before voting.
For a board election, the relevant standard is not whether the institution posts often. It is whether a member can trace a right from the constitution to the register, from the register to platform access, from platform access to a protected ballot, and from the tally to directors taking office. Every exception should have an owner, a reason, a time limit, and an independent review route. That is transparency of control.
The 2022 election material could not cure structural board arithmetic
AFRINIC's archived 2022 election process demonstrates that detailed election calendars can coexist with a structural breakdown. The published material allocated tasks to the board, Nomination Committee, Election Committee, staff, and members. It set dates for nomination, review, proxy registration, electronic voting, the Annual General Members' Meeting, and announcement of results. It was more than a vague promise.
But the plan depended on functioning organs and contested assumptions about vacancies, terms, and the number of seats open for election. Board minutes from early 2022 show debate over how many elections were needed to restore staggered terms and continuity. A published schedule could not resolve a disagreement about the board's own composition if the relevant legal and constitutional interpretation remained contested.
This episode matters because it separates procedural visibility from institutional capacity. A timetable can assign “Board” as the responsible actor. It cannot make the board quorate. A nomination page can invite candidates. It cannot ensure that enough lawful appointments will exist after the meeting. A result notice can state vote totals. It cannot cure a seat term or appointment process that a court later finds defective.
The lesson is not that calendars are useless. They coordinate volunteers, candidates, members, vendors, and staff. They expose lateness. They make selective deviation easier to detect. The lesson is that a calendar must be derived from a verified authority map. Before the first public date, the institution should know every seat's legal status, the source of power for each appointment step, the quorum consequences of each possible outcome, and the fallback if an organ cannot act.
In AFRINIC's case, publication helped observers see the mismatch. It did not repair the board arithmetic. That required legal clarification and, eventually, a court-backed restoration route operating in the absence of ordinary directors.
Receivership solved the convenor problem, not the electorate
The receiver's appointment was a decisive institutional intervention because it answered one question that transparency could not: who could lawfully preserve the company and organise a route back when ordinary board authority was unavailable? It did not answer every downstream question.
The receiver still had to work with AFRINIC's constitution, membership records, staff, committees, vendors, and court directions. Member eligibility depended on more than a public name. It could involve the class and date of membership, completion of formalities, payment status, legal identity, and corporate authority to designate a voter. Candidate eligibility brought a separate set of judgments. Election administration required access to private documents that could not responsibly be posted in full.
This made the temporary custodian unusually powerful. The person charged with ending the emergency also influenced the machinery that would identify the electorate and produce the directors who ended it. Public rules and independent roles were therefore essential conflict controls. Yet role labels alone were not enough. If the same underlying data, staff, or discretionary exception passed through several named committees, formal separation could conceal practical dependence.
A robust design needed at least five distinct controls. The secretariat could maintain member evidence but should not be final judge of a contested exclusion. The receiver could establish the restoration framework but should not choose outcomes through undocumented exceptions. The election bodies could administer published rules but needed an independent route for challenges. The voting provider could authenticate and count but should not create the eligible roll. An external reviewer could test reconciliation and certify the bounded propositions actually examined.
Transparency supports each separation by making mandate, appointment, conflict declarations, decision criteria, and aggregate outcomes visible. It still does not create independence. Independence depends on appointment terms, access to evidence, ability to disagree, protection from removal, and a report that states what was tested. A committee biography page proves identity, not institutional distance.
The receiver model was necessary because the company needed a temporary source of executable authority. The same necessity made it limited public evidence as a complete legitimacy theory. AFRINIC's members remained the principals for electing directors under the constitution. The recovery design had to move authority back to them through a verifiable chain rather than asking them to trust the visibility of the receiver's work.
June 2025 was heavily documented and still failed
The first 2025 election is the strongest evidence against equating transparency with reconstitution. AFRINIC published an election date of 23 June, identified committee and service-provider roles, released guidelines, answered frequently asked questions, described voting channels, announced candidate information, and issued a series of communiques as court proceedings affected the timetable. The process was visible in considerable detail.
On 13 June, the Supreme Court issued an interim order restraining the planned electronic and in-person election dates and directing that the election proceed according to AFRINIC's bylaws and the earlier court order. The order made the matter returnable on 30 June. Subsequent communications described uncertainty, continuance, extensions, suspension, and ultimately annulment.
The transparency record therefore captured activity without guaranteeing convergence. Different documents could be individually clear while the combined authority state remained unstable. A member reading one communique needed to know whether it had been superseded by another. Candidates needed certainty about which deadline controlled. Election administrators needed a rule for preserving already-submitted votes when a court changed the permissible path. Courts needed a record of what the bodies had actually done, not only what their published plans said they would do.
The critical breakdown came when concerns about voter documentation reached the result. The first design allowed several routes to representation, including proxies and powers of attorney. Multiple channels can widen access, but they create different proof entities and different failure modes. A member's underlying right, the corporate authority of a signatory, the authority of a representative, platform credentials, and the secret ballot are separate things. A defect in one does not automatically contaminate the others.
AFRINIC's public account did not establish the full perimeter of the concern before the whole process was annulled. Again, that does not settle the merits of the receiver's decision. It establishes a governance fact: extensive publication did not supply an agreed materiality test or a proportionate remedial ladder when the hardest dispute arrived.
A recoverable election would define those steps in advance. It would preserve the questioned document, notify the affected member, suspend only the relevant credential where safe, obtain independent review, determine whether the issue affected authorisation or vote content, and choose among correction, exclusion, partial recount, rerun of a contest, or whole-election annulment according to a stated threshold. It would publish reasons at the highest level consistent with privacy and investigation integrity.
The absence of that visible ladder turned transparency into narration. The public could follow the collapse. It could not use the published rules to prevent it.
Annulment exposed the difference between reasons and slogans
The receiver invoked transparency, fairness, and the best interests of AFRINIC when annulling the June election. These are legitimate governance values. They are not self-applying rules. Each can support opposite remedies unless connected to evidence and proportionality.
Transparency may favour disclosure and a targeted investigation so valid votes are preserved. It may also favour a complete rerun when the evidence base is too compromised to certify. Fairness may protect the voters who complied with every rule. It may also protect excluded members who faced a tainted authority channel. The best interests of the company may demand speed because the board vacancy is dangerous. They may demand caution because a disputed board could deepen the crisis.
A reasoned decision resolves those tensions. It identifies the known fact, the unresolved fact, the affected right, the available alternatives, the harm of delay, the harm of finality, and the legal basis for the selected remedy. It need not expose a police file or personal document. It must give enough structure to show that the institution did more than attach respected words to a discretionary choice.
The 15 July notice improved the public account by acknowledging uncertainty. It said the investigation had not reached final conclusions and the receiver could not formally report the extent of the identified irregularities. That candour matters. It also makes clear that the public case for whole-election annulment was not a completed factual adjudication.
This is the point at which more announcements reach diminishing returns. Repeating a commitment to transparency does not answer how many authorisations were questioned, whether any ballot outcome depended on them, which alternative remedies were examined, or why they were rejected. A public institution gains trust by revealing the logic of restraint, not by multiplying value statements.
The proper measure is decision usefulness. Does the disclosure let an affected member understand the status of its right? Does it let a reviewer identify the applicable rule? Does it preserve the possibility of correction? Does it distinguish allegation from finding? Does it show who authorised the remedy? If not, the communication may be visible while the power remains opaque.
A deadline mattered because it had a failure consequence
After the June election was annulled, the receiver returned to court and reported an extension requiring new elections by 30 September 2025. This date was more than another item on a website. It was connected to the receiver's judicial mandate and therefore constrained the temporary arrangement.
The distinction is subtle but decisive. An institutional timetable usually says what administrators intend to do. A court deadline states what a competent authority requires. The first can be revised internally, sometimes without a meaningful consequence. The second can support directions, enforcement, replacement, or other judicial response if ignored. That changes incentives.
The September deadline did not guarantee a good election. A rigid date can encourage rushed design, compress challenges, or turn judicial urgency into administrative inflexibility. It did, however, supply what earlier transparency lacked: an external completion constraint tied to the temporary office-holder's mandate.
A sound recovery plan would pair such a final date with intermediate milestones and explicit failure triggers. The member population should be reconciled by one date. Eligibility notices should issue by another. Candidate vetting, provisional register publication, challenges, final certification, platform loading, voting, results, and handover should each have an owner. If a milestone fails, the plan should state who decides whether to cure, extend, replace a provider, or seek directions.
AFRINIC's replacement timetable supplied many of these dates. The remaining governance test was whether exceptions were logged and remedies were timely. A deadline can force completion, but it can also hide unresolved error if success is defined only as announcing winners before midnight.
The significance of 30 September is therefore not that courts should manage Internet registries by calendar. It is that temporary authority needs an expiry condition that cannot be rewritten solely by the temporary authority. Transparency reveals the condition. Enforceability makes it real.
The replacement election reduced ambiguity rather than merely increasing disclosure
The September design changed the control structure. It used one directly designated voter for each eligible Resource Member, prohibited proxies and powers of attorney, required executive authorisation, employed identity verification and biometric activation, published provisional and final register stages, and used an online voting period. The rules also named election bodies and trustees and described a result process.
These controls did not make error impossible. Direct designation can still fail if the underlying corporate contact is stale or the executive authority is misunderstood. Biometrics can test whether a human matches an identity document but cannot prove that the company validly empowered the human. A public register can reveal omissions to members who see it but cannot prove that every eligible organisation received notice. A secure platform can faithfully count the wrong electorate.
The improvement was that the design removed one layer of representational indirection and made the remaining chain easier to audit. Instead of accepting several document types at the point of voting, the system sought a fresh designation from an eligible member, verified a single human, published register stages, and then activated that voter. Fewer paths meant fewer incompatible evidence rules.
The published statistics added a useful participation funnel: 581 total voters, 548 completed biometric registrations, 33 incomplete registrations, and 484 votes cast. Those figures do not supply the full member denominator, but they let a reviewer see attrition after the reported voter population was formed. The register page also described four late confirmations while stating that no new applications were accepted after the official deadline. That distinction should be backed by a versioned exception record, yet its publication gave members a specific event to examine rather than a generic assurance.
Most importantly, the election finished. Eight winners were announced in September 2025. A later joint communication from the board and receiver stated that directors were in place and that the receiver had applied for termination of receivership on 8 October. Those acts changed the institutional position in a way that months of earlier updates had not.
Transparency helped make the replacement design legible. Reconstitution occurred because the authority chain produced a completed election and directors took office.
External supervision added competence, not automatic legitimacy
On 5 September 2025, a court order directed that the election proceed in the presence of the Electoral Commissioner of Mauritius, with oversight and supervision aimed at a free and fair process. AFRINIC then announced the Commissioner's involvement and the final voting window.
This intervention addressed a real weakness. A national electoral office can bring experience in ballot controls, observation, chain of custody, and procedural fairness that a registry does not necessarily possess. Its presence also reduces the appearance that the receiver and AFRINIC staff certified their own route out of receivership.
But external supervision is often overread. An observer can examine only the scope, evidence, and access provided. Presence at opening, closing, or counting does not establish the completeness of the underlying member population. Observation of biometric activation does not decide corporate authority. Supervision of the platform does not determine whether a candidate was lawfully excluded. The public record reviewed for the election did not include a detailed permanent report defining every proposition the Electoral Commissioner tested and certified.
Transparency should therefore attach a certification matrix to external review. It should state whether the reviewer examined the legal member denominator, eligibility calculations, individual challenges, voter-designation evidence, platform configuration, access logs, ballot secrecy, tally integrity, exceptions, and result publication. For each area, the report should state the method and any limitation.
This is not distrust of the Electoral Commissioner. It is protection against mandate inflation. Institutions often cite an independent person's presence as if it validated every layer. A bounded report is stronger because it prevents later actors from using the observer's reputation beyond the work performed.
The court order and electoral supervision were executable safeguards. Their authority and competence mattered independently of publicity. Publication made it possible for members to know that they existed and later ask what they covered. Once again, transparency served accountability; it did not generate legitimacy by itself.
Result publication must reconcile the promise and the permanent record
The replacement guidelines said that the chair of the Nomination Committee would announce results including the total votes for each candidate. The official result announcement reviewed in the public record listed the eight successful candidates. It did not display the per-candidate totals in that permanent notice.
Totals may have been stated in a live event or held in another record. The analytical point is narrower: a promise of public evidence should be satisfied in a durable, linked form. A livestream is not an archive unless retained and indexed. An oral statement is not a usable audit record unless transcribed or attached. A result page should allow a member to reconcile the declared winner with the published tally without searching across temporary channels.
This matters particularly after an annulled first election. The replacement process needed not merely a result but an evidence package capable of surviving litigation, staff changes, receiver discharge, and institutional memory loss. That package should include the certified register version, platform population, opening and closing records, cast-ballot count, spoiled or rejected ballot treatment, per-contest totals, tie handling, observer certification, exceptions, and challenge disposition.
Not all of it must be public at individual level. Ballot secrecy and data protection are non-negotiable. Aggregate and cryptographic evidence can preserve assurance without exposing choices. Sensitive authorisation documents can be retained under controlled access for courts and reviewers.
The board's legal existence does not depend on pleasing every observer. But its long-term authority benefits from a replayable permanent record, especially where the preceding years were defined by arguments over who could act for the company. Transparency is strongest when it follows the evidence lifecycle from rule to archive rather than ending with a celebratory announcement.
Electing directors was a state change; institutional recovery was a sequence
The September result reconstituted a board, but AFRINIC's October joint communication showed that normalisation remained unfinished. The receiver had applied for formal discharge and was still supporting the institution pending the court's decision. The new board was reviewing membership and allocation issues, litigation, financial reports, audits, the rejected 2021 financial statements, policy structures, governance committees, and bylaws reform. Staff had operated without a substantive chief executive for nearly three years.
That list is evidence of institutional honesty: the board's arrival did not erase accumulated liabilities. It also warns against using “board restored” as a complete recovery metric. A quorate organ can exist while authority remains shared with a receiver, accounts remain delayed, executive leadership remains vacant, legal cases continue, and operational decisions await review.
The board later established an interim management committee and a CEO search committee, with receiver consent, and identified six months or appointment of a new chief executive as the interim committee's outer condition. Those actions moved recovery forward. They also showed that governance after election required defined work, resources, dependencies, and reporting.
Transparency is useful at this stage only if it measures completion. Announcing a committee is an input. A recovery report should state its mandate, start date, deliverables, dependencies, decisions, unresolved risks, and end condition. Publishing an audit appointment is an input. Members need filing status, findings, remediation ownership, and the date on which ordinary approval returns. Announcing community engagement is an input. Members need evidence that questions altered decisions or received reasons.
This distinction protects the elected board as much as it protects members. Directors should not inherit an impossible promise that election day cured years of institutional debt. A staged recovery ledger lets them show progress without claiming normality prematurely.
Members needed rights, not an audience seat
Throughout the crisis, “members” could be invoked as recipients of updates, voters, customers, resource holders, or components of a broad community. Those positions are not identical. Transparency becomes weak when it broadcasts to members without giving them a usable right.
A member-level right starts with notice specific enough to act on. Before an election, each Resource Member should receive its legal name, membership class, eligibility state, good-standing calculation, designated contact, missing evidence, deadline, and challenge route. A later change should generate a receipt showing old state, new state, reason, decision-maker, and register version.
The challenge must be decided before the vote becomes irreversible. An appeal delivered after directors take office may clarify precedent, but it cannot return the lost ballot. Emergency timing therefore requires a reviewer with authority to correct, provisionally admit, segregate, or preserve the disputed right. Courts remain available for serious disputes, but ordinary members should not need full Mauritian litigation to correct a billing or designation error.
Collectively, members need denominator information. How many legal Resource Members existed at the cutoff? How many were assessed? How many were eligible, notified, designated, verified, activated, challenged, corrected, and able to vote? A statement that 484 ballots were cast is useful but does not answer how broad the opportunity was.
This is the accountability difference between a transparent audience and a principal. An audience receives information. A principal can inspect the basis of a decision, contest it, require reasons, and ultimately replace the decision-makers under valid rules.
AFRINIC's board vacuum persisted in part because the company lacked a dependable, uncontested route for principals to reassert ordinary authority. The replacement election made that route more concrete. Future resilience should preserve it continuously rather than rebuilding it under court pressure.
Exit rights are the structural test of registry accountability
Even a well-run election cannot carry the entire accountability burden for a registry. Most AFRINIC Resource Members are network operators and organisations focused on service, customers, security, and investment. They may vote rarely, lack time for procedural politics, or reasonably conclude that one ballot has little effect. A governance system that depends on constant participation by every operator will eventually be dominated by those with lower participation costs.
Transparency can expose that imbalance, but it cannot discipline a monopoly if affected networks have no practical exit. Publication of poor performance does not create an alternative service. A public apology does not compensate an operator for delayed registration, disputed resource rights, or an unsafe policy decision. Election rights are necessary internal accountability; portability is external accountability.
This is where the positive direction advanced by the Number Resource Society is relevant. Resource holders need enforceable continuity, auditability, collective protection, and the ability to move registry relationships without renumbering or destabilising customers. Portability would not eliminate the need for unique records or dispute controls. It would prevent continuity from being equated with permanent dependence on one regional corporation.
The AFRINIC crisis showed why both mechanisms matter. Members needed a lawful election to restore the company's board. Operators also needed their networks and customers protected while the corporate route failed. If the only remedy for registry failure is a multi-year court-and-election struggle in the registry's host jurisdiction, transparency describes lock-in more clearly without removing it.
Exit must be engineered carefully. Registry records, RPKI, reverse DNS, transfer history, dispute flags, and contractual obligations cannot be split across incompatible authorities. A portable model therefore needs common technical invariants, authenticated transfer, independent adjudication, and a clear last-verified state. It should move the registration service without making duplicate claims possible.
The principle is simple: voice and exit reinforce each other. Members with meaningful exit are harder to ignore. Registries that can lose service relationships have a structural incentive to make transparency useful, remedies timely, and governance competent.
A transparency architecture for the next vacancy
AFRINIC should not wait for another loss of quorum to reconstruct the evidence needed for recovery. The institution can maintain a standing governance-continuity architecture with five linked records.
First is an authority ledger. It identifies every corporate organ, seat, term, vacancy, delegation, quorum condition, temporary office, court direction, and power boundary. Each material act can then name its authority source. This avoids asking readers to infer power from the letterhead.
Second is a membership assurance record. It separates legal membership, service account, fee status, election eligibility, designation, platform activation, and ballot participation. Changes are versioned and privately receipted to the member. Aggregate reconciliation is public.
Third is an election evidence schedule. Before nominations open, it states accepted proof, contradiction rules, reviewer roles, conflict safeguards, exception logging, platform tests, observer scope, and retention. The remedial ladder addresses a disputed document, voter, contest, or system without assuming that every defect requires total annulment.
Fourth is a recovery milestone map. It gives each task an owner, budget, dependency, due date, evidence of completion, and failure trigger. Court dates appear as external constraints, not as substitutes for operational work. The map continues through receiver discharge, credential handover, audits, executive appointment, committee restoration, and member reporting.
Fifth is a continuity and portability layer. Critical registration services have tested backup, data custody, authority boundaries, member notice, correction rights, and a path to temporary or permanent transfer if the corporation cannot perform. Network and customer continuity outrank preservation of discretionary institutional power.
None of these records requires publication of private identity files, security-sensitive keys, legal advice, or ballot choices. Transparency can be layered: public rules and aggregates, confidential member receipts, controlled reviewer access, and sealed judicial evidence where necessary.
The result is not maximal disclosure. It is sufficient evidence for each person to exercise a right and each reviewer to test a claim.
What the record proves, and what it does not
The public record supports a firm sequence. AFRINIC lost a functioning board and later had no directors in ordinary office. Courts and receivership supplied temporary preservation authority and a route to elections. The June 2025 election was extensively announced and documented but was suspended and annulled after concerns about voter authority documents. The receiver stated that an investigation remained incomplete and that the extent of irregularities could not yet be formally reported. A court-backed deadline required a new election by the end of September.
The replacement design narrowed voter representation, published register stages, added identity controls and later operated under Electoral Commissioner supervision. Results were announced for eight board seats, and a later joint statement said a board was in place while the receiver sought discharge.
The record does not prove that every June authority document was valid or invalid. It does not establish that the whole-election remedy was legally unnecessary. It does not independently certify every September eligibility decision, biometric result, platform control, or ballot. It does not resolve litigation filed after the election or determine the final legality of every act by the restored board. Those matters require the full evidence and competent adjudication.
The article's conclusion does not depend on taking sides in those disputes. Transparency could not reconstitute the board because transparency is not a legal actor, electoral body, member vote, or remedy. It made failures and proposed solutions visible. The board returned only when lawful temporary authority, a bounded timetable, an executable election design, member participation, counting, result declaration, and handover converged.
That distinction should guide the next reform. AFRINIC does not need fewer notices. It needs each notice to point to a right, a responsible actor, an evidentiary record, a deadline, and a consequence. Members should be able to move from information to correction. Courts should be able to move from a return date to measurable compliance. Operators should be able to preserve continuity without being trapped by institutional failure.
The test of transparency is not whether the institution has said enough. It is whether the disclosed system can change a wrong outcome before the people bearing the loss are forced to absorb it.
Sources and analytical limits
The AFRINIC Constitution 2020 supplies the formal company, member, board, nomination, election, and voting framework. It defines the intended distribution of authority but does not prove that a disputed decision complied with the constitution.
The Supreme Court judgment in Benjamin Adzenyamebeye Eshun v African Network Information Centre (AfriNIC) Ltd, 2023 SCJ 63, the 12 September 2023 receiver order, and the Court of Civil Appeal judgment 2024 SCJ 473 support the board-quorum, authority, receivership, and restoration chronology. They do not certify later election execution.
AFRINIC's archived 2022 election process and 2022 board minutes establish the published calendar, assigned responsibilities, and contemporary debate about seat terms and continuity. They do not decide every legal consequence of the 2022 election.
The receiver's 21 April 2025 election communique, the June election guidelines, and the June FAQ establish the first election's announced design. The 13 June 2025 interim order establishes the injunction and directions recorded in that order.
The receiver's annulment communique, 30 June update, and 15 July notice support the attributed reasons, continuing investigation, absence of final conclusions, replacement-election decision, and 30 September deadline. They are not treated as findings that any named person committed fraud or that every vote was affected.
The replacement election guidelines, designation process, voter register, and election statistics establish the announced controls and reported aggregate participation. They do not independently prove the correctness of each underlying membership or voter decision.
The 5 September 2025 order, 6 September announcement, result announcement, and 13 October joint statement support the external-supervision, completed-election, elected-board, and handover chronology. The public material reviewed does not contain the complete private member register, disputed authority documents, police file, voting-system configuration, secret ballots, all challenge decisions, a comprehensive Electoral Commissioner report, sealed court evidence, or final determination of every later election challenge. Conclusions are therefore limited to institutional design and the public evidence chain.