Summary

  • The Internet Society announced on 31 July that Ramanou Biaou, Caleb Ogundele, Sharon Kent and Brian Trammell had been seated for three-year terms at its 25–26 July Annual General Meeting.
  • One challenge was filed after the chapter and organization-member election results were announced on 31 March. The Board rejected it unanimously on 8 April after receiving background information and a review of associated audit results.
  • The minutes and final-results notice do not identify the question raised, the review standard, the audit finding or the reason for rejection. They do not show that the outcome was defective either.
  • The current selection procedure requires a private final decision to the challenger, not a public reasoned notice. That design can protect sensitive information while still disclosing a non-identifying account of the issue, evidence, finding and remedy.

The result is settled; the review record is not complete

The July transition closed the practical part of the 2026 cycle. Ramanou Biaou and Caleb Ogundele took the two chapter-elected seats, and Sharon Kent took the organization-member seat. Brian Trammell joined through a separate route: the Internet Architecture Board selected him for the seat associated with the Internet Engineering Task Force under the process documented in RFC 3677. All four began three-year terms at the Annual General Meeting held on 25 and 26 July. The Board also elected Brian Haberman as chair.

That distinction matters. The public record does not say what the challenge concerned. It cannot fairly be attached to any trustee, constituency or ballot, and it should not be allowed to cast a generalized shadow over all four appointments. Nor does the available evidence establish that the result was wrong. The Internet Society says the Board reviewed the challenge, rejected it and allowed the results announced on 31 March to stand.

The accountability question is narrower: what did the review establish?

The answer matters because the Internet Society describes its Board as the body that provides strategic direction and oversees operations and implementation. Chapters elect four of the possible fifteen voting trustees in the steady-state design; organization members elect four; the IETF appoints four; and the Board may appoint up to three itself. Selection is therefore not ceremonial. It determines who exercises governance over the Society and also serves on the Internet Society Foundation board.

Six minutes, one audit and no public finding

The Board's 8 April videoconference lasted from 15:55 to 16:01 UTC. It was classified as open to observers, although no open-forum session was scheduled. No trustee declared a conflict.

The minutes record two substantive briefings. First, the Elections Committee chair presented the committee's report and received no questions. Then President and CEO Sally Wentworth briefed trustees on the challenge, supplied background information and reviewed “associated audit results.” The chair thanked her and staff for the background and audit findings.

The resolution that followed contains only the result. It states that the Board had received a challenge, rejects it and authorizes the President and CEO to advise the challenge author. Barry Leiba moved the resolution; Brian Haberman seconded it. Every trustee present voted in favor.

That is useful evidence. It establishes that the challenge reached the Board, that an audit existed, that the trustees received information beyond the complaint itself, that no conflict was declared and that the decision was unanimous. The 9 April final-results notice adds the filing deadline and confirms that the original results stood.

But neither document says whether the challenge alleged a ballot-counting error, voter eligibility problem, campaign-conduct breach, process departure, technical failure or something else. The audit's scope and conclusion are absent. So are the clause applied, the standard for disturbing a certified result, the reason the evidence failed or succeeded on any point, and any corrective action short of changing the winners.

The omission creates two symmetrical risks. Readers who distrust the process can imagine that a serious defect was dismissed. Readers who trust it can assume that the complaint was frivolous. The public record supports neither conclusion.

The procedure asks for reasons in, but not reasons out

The current Procedures for Selecting Trustees give a challenge seven days after the Elections Committee's result announcement. It must be addressed to the President and include the reasons for the challenge. The President, after consulting the chairs of the Nominations and Elections Committees and the Board's trustees, must advise the filer of the Board's final decision within seven days after the challenge period ends.

The rule defines a recipient, deadline, consultation chain and final decision. It does not publish an eligibility rule for who may file, enumerate permissible grounds, state a burden or standard of proof, require the preservation of a particular record, specify recusals, distinguish harmless from outcome-changing error, or prescribe any public decision notice. The filer must provide reasons; the institution is required only to communicate a result back to that filer.

Private communication may have contained more detail in 2026. The public sources do not establish what the challenger received, so the absence of public reasons should not be converted into a claim that no reasons existed at all. Yet a private explanation cannot perform the same institutional function as a public one. It cannot tell future candidates which grounds are reviewable, show electors what the audit checked, or demonstrate that a procedural weakness was corrected even when the winning slate remained unchanged.

Finality and reviewability answer different questions. Finality says who gets the last word and when the election can close. Reviewability shows how the decision-maker moved from allegation and evidence to conclusion. A system can have an authorized final decision and still leave too little public material for anyone outside the room to learn from it.

The June amendment does not close the evidentiary gap

On 10 June, the Board considered revised Procedures for Selecting Trustees in a fifteen-minute videoconference. The published resolution says the Governance Committee recommended changes to align the procedures with current practice, simplify their language and organization, and address previous concerns raised by trustees. The current procedure page identifies itself as modified by Board Resolution 2026-13.

The timing is notable but not proof of causation. The public agenda does not say the April challenge prompted the amendment. It does not identify which trustee concerns were addressed, and the public page does not supply a redline alongside the resolution. An accountable analysis must therefore keep the two facts separate: a challenge was rejected in April; the wider selection procedures were amended in June.

What can be assessed is the rule now in force. Its challenge clause still centers on a submission to the President, consultation with committee chairs and trustees, private advice to the filer and a final Board decision. It does not require a non-identifying public statement of grounds and reasons.

This is a missed opportunity, not evidence of a concealed defect. A revised procedure could preserve the existing deadlines and authority while adding a simple publication rule. Once a challenge is decided, the Society could issue a notice stating the category of issue, governing provision, material reviewed, audit scope, standard applied, finding, vote and recusals, disposition, and any process correction. Details that could identify a person or reveal a ballot could be withheld with a stated reason.

Confidentiality is compatible with a reasoned notice

Election challenges often contain material that should not be public: names of voters, security logs, candidate communications, personal allegations or evidence that could expose a secret ballot. A serious transparency proposal must not make filing a challenge a route for releasing that material.

The relevant distinction is between source disclosure and decision explanation. A court can publish a redacted judgment. An auditor can report a control finding without exposing raw logs. An election authority can classify an issue and state why it did or did not affect an outcome without identifying the complainant. The Internet Society can do the same.

A compact notice need not quote the challenge. It could say, for example, that the Board reviewed an allegation concerning voter eligibility under a named section; examined the certified voter list and system audit; applied an outcome-materiality standard; found a specified number of records affected or none; recorded any recusals; and either confirmed the result or ordered a remedy. That structure would be meaningful even if every identity and ballot remained sealed.

Lu Heng's distinction between participation and mandate is useful here as an evidence discipline. Chapter and organization-member ballots are real internal acts; they are not made more legitimate by simply invoking “community.” Their force comes from the Society's bylaws, procedures and authorized constituencies. When the institution also offers a challenge process, the public record should show what that process protects. Otherwise participation is counted, authority is exercised and finality is announced, but the bridge between them remains private.

Sources