Summary

  • AFRINIC’s own minutes record that members elected Viv Padayatchy, coded MU, as Primary and Kenny Yiptong, also coded MU, as Alternate for Indian Ocean Board Seat 6 on 27 April 2005. That closes the identity of the institutional act, the seat label and the announced officeholders, but not the integrity of the unseen ballot process.
  • The only published 19–11 Seat 6 tally belongs to the founding election of 24 May 2004. The AFRINIC-2 record does not reproduce a complete 2005 candidate slate, vote totals, turnout, voter register, representative-authorisation ledger, count certificate or challenge record.
  • The outcome passes a narrow geographic-classification test because both officeholders carried Mauritius codes in a seat labelled Indian Ocean. It does not prove broad participation across Indian Ocean economies, substantive regional consent or a reproducible mandate from member companies.
  • The strongest contrary account is plausible: a young registry, newly recognised and completing a planned staggered transition, may rationally have retained an experienced pair that members had selected eleven months earlier. The absence of a detailed public record does not establish fraud, illegality, capture or even that the election was uncontested.
  • Seat 6 allocated bounded control inside a Mauritian membership company. It created no sovereignty, political ownership of regional number resources or election-derived power to punish resource holders or impair running networks outside valid contract, applicable law, due process and competent authority.

Analysis

A result line, and the audit it cannot complete

On 27 April 2005, during the member AGM at AFRINIC-2 in Maputo, the African Network Information Centre recorded that two Board seats had fallen vacant and that members were required to elect replacements. Seat 5 was assigned to Central Africa. Seat 6 was assigned to the Indian Ocean sub-region. For Seat 6, the published result named Viv Padayatchy of Mauritius as Primary Director and Kenny Yiptong of Mauritius as Alternate Director.

That short result line carries genuine evidential weight. It identifies the institution, meeting, date, constituency, corporate offices and people whom AFRINIC recognised as holding them. It is the contemporaneous institutional record of an act AFRINIC called a member election. There is no reason to dilute that basic conclusion merely because the record is concise.

But an official result and a reproducible election are different evidential objects. The public AFRINIC-2 record does not state how many member organisations were eligible, which organisations actually participated, who was authorised to act for each one, how many ballots were issued or returned, what voting weight applied, whether proxies were used, how invalid or abstaining ballots were handled, or whether any objection or recount occurred. It does not identify a complete Seat 6 candidate slate. It supplies no candidate totals and no margin.

It does not explain from the contemporaneous rule text how the Primary and Alternate ranks were obtained.

The correct finding is therefore neither dismissal nor certification. AFRINIC’s minutes are strong evidence of what AFRINIC did, recorded and claimed. They are not an independent examination of the underlying register, authorisations, ballots and count. This distinction applies equally to supportive institutional material: an official document may establish an institution’s representation or decision without independently proving ballot integrity, legal validity, authorised company representation or the truth of every assertion in that document.

The distinction matters because the two visible promises behind Seat 6 are not the same. One was geographic: an Indian Ocean office should be filled within the declared regional classification. The other was corporate: the organisations holding membership rights, rather than an undefined audience around the meeting, should be the principals behind the vote. The first promise is visible in the result. The second requires a chain of records that the surviving public account does not expose.

What belongs to 2004, and only to 2004

The previous year supplies the essential comparator, but it must be kept in its own column. At AFRINIC-1 on 24 May 2004, five of the seven then Full Members present unanimously extended voting rights beyond that small group to three stated categories of organisations: formal AFRINIC supporters; established local Internet registries or Internet service providers with registry or membership identifiers; and organisations that completed membership and paid the membership-only fee before the AGM. The report then counted 32 eligible persons and said 30 expressed a vote.

In that 2004 founding contest for the Indian Ocean position, the tally was 19 for Padayatchy and 11 for Yiptong. The records vary the forms of their names: the tally uses “Vivi Padayatchy” and “Kenneth Yiptong,” while the resulting Board list and the 2005 account use “Viv Padayatchy” and “Kenny Yiptong.” The founding report assigned the higher result to the Primary role and the other result to the Alternate role.

Those quantities support calculations about 2004 alone. Thirty votes expressed by 32 eligible persons correspond to 93.75 per cent recorded participation among those eligible persons, with two not expressing a vote. Within the recorded Seat 6 total, 19 and 11 amount to approximately 63.33 per cent and 36.67 per cent, separated by eight votes or about 26.67 percentage points. None of those figures describes the 2005 renewal. They cannot be carried forward simply because the names and ranks were carried forward.

The year boundary is absolute. The public 2005 AFRINIC-2 record does not reproduce a complete candidate slate or any Seat 6 vote totals. It does not establish that the two incumbents faced one another again, faced other candidates, stood alone, were returned without opposition or received the same support. A later discovery could change the audit: a genuine 2005 count sheet might show another 19–11 result. Until such evidence is authenticated and the package is resealed, however, 19–11 belongs only to 2004.

Meeting attendance offers no substitute denominator. AFRINIC-2 drew more than 100 participants, including members, directors, other registry representatives, government officials, operators and other attendees. That mixed meeting population was not automatically the corporate electorate. Dividing an unknown ballot figure by the meeting headcount would manufacture a turnout rate from unlike categories. Participation in a technical or governance meeting can provide expertise and observation; it does not, by itself, create authority to cast a member organisation’s vote.

The founding sequence also explains why the same seats returned to election so quickly. The 2004 design gave Seats 5 and 6 one-year startup terms, while other Board pairs received longer initial terms, so the pairs could enter a staggered renewal cycle. The 2005 vacancy language should therefore not be read as evidence of removal, misconduct or an unexpected rupture. It describes the scheduled expiry of that transitional pair in the sequence reported at the founding meeting.

This chronology makes personnel continuity measurable without turning it into a vote claim. Both published Seat 6 officeholders in 2004 appeared again in the same roles in 2005: two of two named offices, or complete personnel continuity. The sibling Central Africa seat retained its Primary but changed its Alternate, and the 2005 minutes described the new Alternate there as the only new person on the Board. That contrast shows Seat 6’s distinctive continuity. It says nothing by itself about voter motive, candidate availability, incumbent performance or the intensity of a 2005 contest.

Geography: a pass at the label, an open question about the constituency

The geographic test has a clear, narrow answer. AFRINIC labelled Seat 6 as Indian Ocean. It published both Padayatchy and Yiptong with the country code MU. At the result level, the two named people therefore matched the constituency classification that AFRINIC recorded. The available evidence supports formal Indian Ocean classification.

It does not follow that the result represented the Indian Ocean broadly. Two offices were filled by people carrying one published country code. That is a fact about concentration in the named outcome, not proof of a defective election. The public record does not reveal whether candidates from other Indian Ocean economies were nominated, rejected, withdrawn or defeated. Nor does it show where nominators, seconders, eligible member organisations or voting representatives were distributed.

The distinction between formal origin and substantive representation is not semantic. A regional seat may be useful precisely because it reserves access to Board deliberation for people associated with a service sub-region. It can reduce the risk that all directors emerge from the largest markets or the easiest travel hubs. It can ensure that operating conditions specific to a sub-region have an institutional route to the Board. Those are real governance benefits even when the officeholder is not an instructed delegate.

Yet the label cannot carry more democratic weight than the selection process gives it. To claim broad Indian Ocean representation, an analyst would need more than the MU codes of the winners. Candidate eligibility criteria, the geographic breadth of the eligible membership, the distribution of participation, the authority of the voters and the nomination pathway would all matter. The 2005 record does not supply them. It therefore supports a formal classification, not consent from Indian Ocean operators, users, residents, states or economies as a whole.

Nor does the concentration establish exclusion. The record does not prove that other eligible people were barred, that candidate classification was wrongly applied or that a constitutional diversity requirement was breached. The exact eligibility test—whether based on residence, nationality, domicile, affiliation or another connection—is absent from the authenticated 2005 materials available here. AFRINIC’s printed MU codes show the classification it adopted; they do not reveal the verification file behind it.

The proper representation finding is consequently asymmetric. It is possible to say that the announced outcome fitted AFRINIC’s published Indian Ocean category. It is not possible to say that the result embodied the will of a broad Indian Ocean electorate. It is equally impermissible to say that two Mauritius-coded officeholders made the outcome unlawful merely because the sub-region included other economies. Category compliance, geographic diversity, electorate breadth and legal validity are four separate questions.

This boundary also prevents a corporate constituency from becoming a fictional political people. Seat 6 organised an office within AFRINIC’s Board. It did not appoint sovereign representatives of the Indian Ocean, create a mandate to speak for every network in the sub-region or vest political ownership of number resources in the officeholders. A service region is an administrative map for coordination; it is not a title deed.

The company-member promise and the missing middle

AFRINIC’s March 2005 application for final recognition described the organisation as registered, non-governmental, not-for-profit and membership-based. It placed members above the Board in the control structure and said that members elected Board representatives. Before the renewal, the application also identified Padayatchy and Yiptong as the Indian Ocean Primary and Alternate. These statements provide useful contemporaneous context: AFRINIC represented to ICANN that organisational members were the principals in its governance.

The AFRINIC-2 minutes are consistent with that account. They do not call the wider meeting audience the electorate; they say members were required to elect the replacements. At the declared-process level, that is an important point in favour of the result. The act filled offices in a membership company, not seats in a public legislature or positions awarded by an informal show of hands among all participants.

But a legal entity cannot walk to a ballot box. An organisation acts through a natural person. The integrity chain therefore runs from an eligible member organisation, through a current and duly authorised representative, to one permitted ballot, then through secure counting, ranking and certification to the announced officeholders. The public 2005 record exposes the first claim and the last result, but not the links between them.

For a company-member audit, a list of individuals is not enough unless each individual can be tied to the relevant principal. A conference registration, contact entry, technical role or past association does not necessarily prove current authority. Conversely, the absence of public authorisation documents does not mean valid documents did not exist internally. The honest conclusion is limited to reproducibility: an outside reader cannot reconstruct from the checked record which eligible organisation authorised which human to exercise its corporate right.

The membership denominator is similarly missing. The 2004 report described specific organisation categories and counted eligible persons, but it did not fully expose the organisational denominator underlying those people. The 2005 minutes use the word “members” without publishing the register, good-standing rule, notice record or transition from the temporary 2004 extension to the ordinary electorate. It is therefore not possible to determine company participation, equal voting weight, abstention, duplicate control or whether every eligible organisation had access to the process.

Later constitutional material cannot repair the gap by implication. AFRINIC’s official 2007 constitution describes member election rights, corporate representatives and proxies, six regional Primary-and-Alternate pairs, sequential renewal for the Indian Ocean and Central seats, and a highest-versus-second-highest ranking rule. That is valuable evidence of institutional continuity after the event. It is not the authenticated text governing the April 2005 ballot, and it cannot be silently backdated to supply missing rules.

The 2004 report says a new constitution had been adopted, while the 2005 minutes say the election proceeded according to the bylaws. Contemporary references also point to a February 2005 election process. The exact effective constitution, amendment history and election-process text have not been authenticated in the checked record. Without them, a writer can describe the sequence AFRINIC recorded but cannot deliver a legal opinion that every procedural condition was satisfied.

This leaves the company-member test at a partial pass. The declared principal class was the membership. The outcome filled internal corporate offices. The same pair previously chosen through an organisation-derived founding electorate was returned. But eligibility, authorisation, voting weight, notice, ballots and certification remain publicly unclosed. The record supports what AFRINIC called a member election; it does not reproduce a company-member mandate.

Recognition nineteen days earlier was context, not certification

On 8 April 2005, nineteen elapsed days before the Seat 6 renewal, IANA reported on AFRINIC’s recognition as the regional Internet registry. Its assessment described overall supervision by representatives elected from six identified sub-regions and found AFRINIC’s bottom-up structure acceptable under the recognition framework. ICANN then gave final recognition.

That institutional assessment strengthens the continuity case. It shows that an external coordinating body reviewed the governance design near the time of the election and accepted AFRINIC’s regional and membership-based architecture. It is not trivial evidence and should not be dismissed simply because the later ballot file is incomplete.

Its scope must nevertheless remain exact. IANA did not administer the 27 April member ballot in the located evidence. Its report did not publish the eligible companies, authorised representatives, candidate slate, vote count or challenges for Seat 6. Recognition of the registry function and review of a governance framework do not substitute for a company’s own valid election under its operative instruments and applicable law.

The recognition report therefore proves an institutional assessment and an institutional act: IANA described AFRINIC’s structure, and ICANN recognised the registry. It does not independently establish that every assertion AFRINIC made about its governance was realised in the later election. External approval of a design is evidence about the design, not certification of every future implementation.

The same source discipline applies to AFRINIC’s own application and minutes. They are indispensable because they are the institutional records closest to the act. They should be read as evidence with defined claimants, not converted into omniscient accounts. Official provenance can make a statement authoritative as a record of what an institution said; it does not make every unseen premise independently true.

The continuity defence deserves its full weight

The most convincing defence of the 2005 renewal is straightforward. In May 2004, AFRINIC had published a real founding contest for the Indian Ocean pair. The recorded 19–11 outcome selected Padayatchy as Primary and Yiptong as Alternate. Their unusually short first term was not a sign of failure; it was part of the planned staggering of Board renewals. By March 2005, AFRINIC publicly listed them in those roles while presenting a member-over-Board governance model. In early April, the registry received final recognition after its regional design was reviewed.

Later that month, members renewed the same pair and thus preserved institutional knowledge during an operational transition.

A concise meeting report need not reproduce every ballot paper for the underlying election to have been valid. Internal registers, authority documents and count materials may once have existed even though they are not in the public record checked at the cutoff. Corporate recordkeeping practices in a young institution need not match the later expectations of forensic researchers. Complete continuity can be a rational member choice, particularly when a newly recognised registry is moving from transition to stable operation.

Nothing in the sealed evidence disproves that account. No competent finding in the package calls the election fraudulent, captured, unlawful, discriminatory or a sham. No evidence establishes an unauthorised ballot, insider rotation, manufactured candidacy or improper exclusion. The missing slate does not prove the pair were unopposed, and the repeated ranks do not prove that they competed against one another. Silence about objections does not prove that objections existed or that remedies were denied.

The audit conclusion survives this defence because it is narrower than an accusation. The question is not whether a plausible valid-election narrative can be built; it can. The question is whether the public evidence permits an independent reader to test that narrative against the principals, authorisations, slate and count. It does not. Continuity may have been wise and properly chosen. The surviving record still cannot distinguish a fresh competitive vote from a limited contest, a ranked two-person process or an acclamation-like renewal.

Documents that would strengthen the defence are concrete rather than rhetorical: the authenticated constitution and February 2005 election process; notice sent to members; the eligible organisation register; candidate nominations and qualification records; corporate representative or proxy authorisations; a signed count sheet including abstentions and invalid ballots; and an election-committee certificate with any challenge log. Their value would lie in connecting the company principal to the final result, not in adding more general praise for bottom-up governance.

Why a small corporate election matters to operators

The Board election was internal, but the institution it controlled maintained a scarce-resource ledger. Directors could influence budgets, executive supervision, registry strategy and the conditions under which administrative services were delivered. For operators, uncertainty around Board authority can therefore become governance risk around fees, transfers, registry changes and continuity. The package does not show that this particular 2005 result caused any adverse operator action, so the risk must remain structural rather than alleged harm.

A reserved geographic seat can reduce that risk if it gives a sub-region reliable access to Board deliberation. It can also preserve context during institutional transition. Seat 6’s complete continuity may have helped retain knowledge just as AFRINIC moved beyond final recognition. The risk on the other side is a representation discount: when a regional label is visible but candidate eligibility, member authorisation and the voting denominator are not reviewable, counterparties cannot tell how much substantive consent sits behind the label.

The relevant corporate principal is the organisation that bears fees, operational obligations and possible downside. Its vote should be traceable to a duly authorised human. If that chain is weak, a process can become vulnerable in principle to duplicate representation, stale authority or ballots that do not express the legal entity’s will. These are reasons for record design, not allegations about 2005. No specific fraud or unauthorised vote is established here.

The economic importance of the ledger must not be confused with political power. AFRINIC coordinated unique number-resource records; independent networks performed routing. Registry records, contractual rights, technical use and commercial value occupy related but distinct layers. Influence over an important ledger makes sound corporate governance essential, but scarcity does not transform a private coordinator into a sovereign.

The authority created—and the authority that was not

The Seat 6 election could choose directors of AFRINIC Ltd. and allocate internal Board control under the company’s valid instruments and Mauritius law. A lawfully constituted Board could manage corporate affairs, supervise executives and administer contracts and procedures within its bounded remit. Those powers are real. Calling AFRINIC non-sovereign does not erase corporate obligations or make court-enforceable duties optional.

The election did not confer ownership of African or Indian Ocean number resources. It did not create taxation, policing, confiscation or public-law punishment powers. It did not authorise directors to impair routing, registry continuity, RPKI, reverse DNS or other running-network dependencies merely because a resource holder disagreed with a governance preference. Any binding adverse action would still require a valid contractual basis, applicable law, due process and competent authority.

This is the controlling doctrine boundary: AFRINIC is a number-resource ledger coordinator, not a sovereign. Geography may organise registry service and Board composition, but it cannot manufacture political ownership. Election legitimacy can authorise bounded corporate administration; it cannot turn a regional seat into a mandate over a people, attendance into company authority or an address ledger into an enforcement state.

The boundary also protects the election’s proper importance. If Board authority is exaggerated into sovereignty, every governance dispute becomes a struggle over supposed control of the Internet. If it is dismissed as irrelevant because networks route independently, the corporate levers that affect registry service are ignored. The accurate middle is more demanding: directors matter because they govern an economically significant coordinator, and they remain constrained because that coordinator is private, administrative and legally bounded.

Other stakeholders retain their own roles. Governments, operators, end users, technical experts, ICANN, IANA and other registries may provide evidence, coordination or recognition. None of those roles automatically replaces the member organisation as the corporate elector. Being affected by a registry decision is not the same as holding the company vote, just as holding a company vote is not the same as receiving sovereign authority over everyone affected.

NRS can advocate, research, convene and represent members who explicitly authorise it. It does not operate AFRINIC’s registry, RPKI, directory services, appeals, settlements, elections, custody or continuity. Its analysis may sharpen questions about membership rights, but it cannot fill an absent 2005 ballot ledger by institutional substitution.

A precise conclusion from an incomplete record

The 2005 Seat 6 renewal satisfies the narrowest founding test. AFRINIC’s contemporaneous minutes record an Indian Ocean member election, and the two announced officeholders carried Mauritius codes consistent with that label. The result also displays complete personnel continuity with the founding pair and fits the planned one-year startup term for Seats 5 and 6.

It fails the strongest public-audit test because an independent reader cannot reproduce the 2005 choice. The complete candidate slate, voter register, member-company denominator, representative authority, voting weight, ballot count, ranking mechanics, certification and challenge record are not present in the checked public evidence. That failure is a statement about the surviving record, not proof that the underlying election was invalid or that no internal records existed.

The two founding promises must therefore be reported separately. Formal geographic classification is supported. Broad Indian Ocean representation is not proven. AFRINIC’s assertion that members elected is supported as an institutional act. A reproducible company-member mandate is not proven. The continuity defence remains plausible and material. The 2004 tally remains strictly in 2004.

Above all, neither success nor uncertainty in this corporate election enlarges the nature of the institution. The vote allocated internal control over a number-resource coordinator. It did not award political ownership of the region, sovereign punitive power or a licence to put running networks at risk without valid contract, law, due process and competent authority.