Summary

  • AFRINIC’s 2026 Election Guidelines made the register compiled for the disputed September 2025 Board election the baseline for elections and proposed member resolutions in 2026.
  • Members could confirm, amend or replace a designated voter, but silence by the deadline was deemed confirmation. The public record reviewed here does not say how many designations were affirmatively renewed, changed, challenged or carried forward by default.
  • The active designation route demanded corporate authorisation, identity evidence and AFRINIC verification. The rules do not say that an unchanged, inherited designation passed through those checks again.
  • Five candidates contested three Governance Committee seats. AFRINIC named Daniel K. Nanghaka, Afaf El Maayati and Maud Adjeley Ashong Elliot as elected on 30 June without publishing vote totals, turnout, invalid ballots, trustee identities or a certification record in that announcement.
  • This is not proof that a ballot was false or that the election was unlawful. It is a public audit gap between the member that owns the vote, the person allowed to cast it and the governance result AFRINIC declared.

The consequential decision came before anyone voted

The most important choice in AFRINIC’s 2026 Governance Committee election was not made inside the ballot box. It was made on 4 February, when the persons presented by AFRINIC as its Board approved rules that decided whose earlier authority would survive into the new cycle.

Section 12 of those rules did not begin from a blank voter list. It declared the register assembled for the September 2025 Board election to be the baseline for AFRINIC elections and proposed member resolutions in 2026. Resource Members were offered a period in which to confirm, amend or replace their designated representative. If AFRINIC received none of those actions by the deadline, however, the existing designation was “deemed confirmed” and remained valid for the year.

That mechanism converted absence of a new instruction into continuing delegated authority. It may have saved members from repeating paperwork. It may also have preserved designations their organisations genuinely wanted. But administrative convenience is not the same thing as affirmative proof, particularly when the source election remained contested.

NRS has said the September process was neither lawfully nor conclusively completed and invited organisations to report cases in which their name appeared on the register despite not registering or voting. That is NRS’s stated position and request, not a judicial finding and not evidence against a named representative. AFRINIC, for its part, described the September turnout as unprecedentedly high, acknowledged litigation seeking to invalidate the directors and told members to contact it if they believed they had been falsely represented.

Those rival descriptions make fresh audit evidence more important, not less. A disputed register should not become reliable merely because the next rulebook calls it a baseline.

Silence carried authority forward

A Resource Member’s vote is an organisational asset. The member is the principal; the natural person on the register is a delegate. That distinction is easy to obscure when an election page displays only names and voting status, but it is the centre of the accountability question.

An affirmative renewal would show that the principal had instructed the delegate again. Default confirmation shows only what the rule did when AFRINIC recorded no timely change. It does not reveal whether the notice reached the right executive, whether the member deliberately accepted the existing representative, or how many organisations fell into that category. Nor does the fact pack prove that any particular member failed to receive notice. The missing item is aggregate, privacy-preserving evidence about the process.

AFRINIC’s active 2026 designation guide demonstrates what such evidence can look like for a new or changed representative. It required a letter on company letterhead, an executive signature, government identification, confirmation emails to the executive and registered contacts, and verification by AFRINIC. Those controls create a visible chain from organisation to voter.

The guide does not state that every designation carried over unchanged was put through those steps again. Without a count of affirmative confirmations, amendments, replacements, rejections and defaults, readers cannot tell how much of the 2026 electorate rested on fresh instructions and how much rested on institutional inertia.

The 2025 portal had reported 581 voters, 548 completed biometric registrations and 484 ballots cast by 12 September. Those numbers show scale; they do not validate every underlying designation. The comparable 2026 bridge is what remains absent from the bounded public record.

“Independent” does not map the reporting line

The same rules describe the Election Committee as independent and give it responsibility for voter verification, ballot issuance, collection, counting and tabulation. Yet they also make the committee administratively accountable to the chief executive or people assuming that role. AFRINIC’s Bylaws say the committee consists of staff designated by the CEO.

That combination may be workable, but it is not self-proving independence. Independence is a testable distribution of authority: who appointed the people, who could instruct them, who controlled the records, who resolved exceptions and who certified the outcome.

The trusteeship layer raises a parallel question. Under the guidelines, election trustees comprised the NomCom chair, AFRINIC’s external legal adviser and the CEO—or, in the absence of a CEO, one representative of the Interim Management Committee. They oversaw opening and closing and would settle a tie by drawing a winner. At the freeze for this briefing, AFRINIC’s Team page still showed a blank CEO name cell. That leaves the public rule pointing toward an interim manager unless an appointment not visible in the record changed the position.

No fact here shows misconduct by a committee member, trustee, candidate, adviser or manager. It does show that the same acting-management and legal structure sat close to voter verification, ballot administration and trusteeship. Naming the people who actually filled those roles would therefore be basic audit information.

Winners were named; the certification bridge was not

AFRINIC published five candidates for three Governance Committee places on 17 June. Its mailing-list announcement dated 30 June named Daniel K. Nanghaka, Afaf El Maayati and Maud Adjeley Ashong Elliot as elected.

The notice establishes who AFRINIC declared the winners. In the captured 3,825-byte announcement, it does not state each candidate’s total, the number of eligible voters, ballots issued, turnout, rejected or invalid ballots, challenges, confirmation changes, the voting platform, trustee identities or the signer of a certification record.

Some or all of that information may exist privately, in recordings, meeting papers or pages not located in this bounded search. The claim is limited to the announcement and source set examined. But AFRINIC’s own rules said NomCom should publish final results and statistics and that relevant documentation should be available on the website. The sparse result therefore falls short of the institution’s published transparency promise.

The public timetable had put publication of the final register on 14 June and the GovCom result on 25 June. An indexed community-register page labelled itself 2026 while displaying “Published on 14/06/2025”; that is a metadata defect, not evidence of falsification. The result email arrived on 30 June. At the evidence freeze, the election portal and several register and guideline endpoints returned HTTP 500 in sequential checks. That establishes temporary unavailability only, not concealment.

What a reconstructable result would contain

The remedy is not to publish members’ private identity documents. AFRINIC could preserve confidentiality while disclosing the counts and procedures that make delegated authority reconstructable.

It could state how many organisations received notices, how many deliveries succeeded, how many representatives were affirmatively confirmed, replaced or amended, how many defaults were applied, and how many designations were challenged or rejected. It could name the Election Committee and trustees who served for the vote, identify the platform and any independent auditor, and publish eligible-voter, ballot, candidate-total, invalid-ballot and dispute figures with a signed certification report.

Such a record would not guarantee legitimacy. It would allow members to test it. That distinction matters because governance language—community, consensus, integrity and stewardship—does not transfer authority by itself. Authority travels through an actual chain of principals, delegates, administrators and reviewers.

AFRINIC’s 2026 rules made continuity easier than fresh affirmative proof. Its result tells the public which three candidates were declared elected. It does not yet show the audit bridge from the disputed 2025 designation layer to that 2026 outcome. That is a representation and accountability gap, not a finding that a particular ballot was false.

For the broader economics and legitimacy questions surrounding AFRINIC’s election architecture, readers can consult BTW’s separate research: AFRINIC election legitimacy.

Sources