Summary

  • Article 5 §5.5 of the ICANN Bylaws, in the consolidation amended 3 July 2026, makes an annual analysis of complaints and their resolution a duty of the Office of the Ombudsman, with trends and recommendations to minimise future complaints.
  • The FY25 report that discharges that duty prints closed-case outcome shares reaching 112, with no denominator and no overlap rule; the instruments reviewed here set no reconciliation standard, which is a verification gap rather than a finding of bad faith.

Open page 16 of the Office of the Ombuds' annual report for fiscal year 2025 — a 25-page document covering 1 July 2024 to 30 June 2025 — and the closed-case outcome chart leads with its biggest share. Self-managed, 70 per cent. Process concluded, 13. No further action, 10. Referral, 7. Those four figures make exactly one hundred before anything smaller is counted. The categories beneath them, which record cases declined or withdrawn or abandoned, complaints unsubstantiated, and complaints partly or fully substantiated, sit on top of that hundred, and our own addition of every share printed on the chart reaches 112.

Whether that reflects overlapping categories, a different denominator or a drafting slip is not established by the documents examined. What is established is that the chart, as inspected, prints no denominator and no overlap rule. A reader who wants to check the outcome rates of ICANN's Ombudsman against the cases the office actually handled has nothing on that page to check them against.

The duty behind the page is in the Bylaws. Article 5 §5.5, in the consolidation amended 3 July 2026, requires the office to produce an annual analysis of complaints and their resolution, and calls for trends and recommendations to minimise future complaints. The office is constituted by the Board: §5.1(a) provides a full-time Ombudsman appointed by the Board for a renewable two-year term and requires a three-fourths Board vote for dismissal. §5.3(c) permits declining complaints and excludes internal administration, personnel, Board membership and vendor or supplier matters. §5.4(c) states that contacting the Ombudsman is not notice to ICANN of an action or cause of action. §5.4(d) authorises reports to the Board on particular matters, published by default unless the Ombudsman decides otherwise. In the text of Article 5 as inspected, nothing prescribes how the annual analysis is to be computed.

What an upheld complaint buys is set out in the framework and process approved 7 June 2026, and the ceiling is explicit. A complaint may come out substantiated, partly substantiated or unsubstantiated. An upheld complaint yields written feedback and possibly recommendations. The office cannot unilaterally reverse an ICANN decision. Both parties may comment on preliminary findings, normally within 15 calendar days, and the framework's targets — acknowledgment within two to five working days, most investigations concluded within 60 to 90 days after agreeing scope — are framed as aims rather than guaranteed deadlines.

Then there is the timing, and it is the sharpest version of the problem. The reported period ended 30 June 2025. On our chronology that year ran entirely under the framework approved 26 March 2009, which was superseded on 7 June 2026. The 2009 instrument withheld any power to make, change or set aside a policy, administrative or Board decision, required complaints in writing, and required written notice to a complainant when jurisdiction was declined. The FY25 report was announced on 11 June 2026 — four days after the replacement framework took effect. Its disposition vocabulary belongs to the retired instrument. The current framework's three outcomes are a different list, and the report's categories cannot be mapped onto them from the documents reviewed. That is a mapping gap across two instruments: a structural fact about which rules were in force when, and not an accusation about anyone's bookkeeping.

Page 16 is not the only place where the figures stop short of closing. Page 11 of the same report records three open cases. Its charts read 6 per cent open and 94 per cent closed, with the denominator unstated, so whether those three and that 6 per cent describe one population is not established. Page 11 also gives 213 out-of-scope submissions for the previous year, where the report's own Appendix 1 states 212 in both years.

The FY24 report shows the same pattern in a shorter document. It states 237 submissions, 24 jurisdictional and 213 non-jurisdictional. As of 30 June 2024 it reports 20 jurisdictional complaints resolved, two under investigation and two withdrawn. For non-jurisdictional submissions it lists 49 referrals, 155 self-help responses, three abandonments and two resolutions. Our own addition of those four figures is 209, which leaves four of 213 unaccounted for in that breakdown. The cover ends the period on 30 July 2024 while the executive summary and the complaint-status section both use 30 June 2024. Its page 4 pie chart labels reverse the jurisdictional proportions that its own counts of 24 and 213 imply.

One figure in the series does close, and it is worth putting on the record because it shows what a checkable claim looks like. Twenty-four jurisdictional complaints in FY24 and 36 in-scope cases in FY25 differ by exactly 50 per cent, which is precisely the increase the release announcement claims for the year. The comparison is between an in-scope case count and a jurisdictional complaint count, the two reports use different labels for the two populations, and no reviewed document establishes that the populations are equivalent. But the arithmetic reproduces, which is more than can be said for the outcome chart.

Responsibility for this sits in three places and they should be stated one at a time. The office writes about itself: the annual report is the Ombudsman's account of the Ombudsman's work, and ICANN is the body whose Bylaws constitute the office and to whose Board it reports. The Board appoints the Ombudsman and can remove the holder on a three-fourths vote, so the incentive to read the office's figures closely runs through the Board's own appointee. And the charter, as reviewed, prescribes no denominator convention, no overlap rule and no outside reconciliation of the annual analysis. That is the gap this briefing names: a compliance-and-verification gap, not bad faith. No reviewed document shows that anyone intended to obscure anything.

The fix, if one comes, will be written into an instrument rather than into better prose: a stated denominator, an explicit rule for overlapping categories, and a reconciliation performed by someone other than the author of the analysis. Until something of that kind appears in the reviewed instruments, the honest description of the annual report is a narrative the office writes about itself, and closure in it is not resolution. Self-managed cases require no report-back, so a case counted as closed need not have produced a finding, a remedy or even a reply.

Three limits belong beside those claims. The FY26 report was not obtained in this review; a 13 October 2026 Board Chair blog announces a planned presentation of it covering activities, case data, trends and systemic observations, prints no statistics, and does not establish that the report has been published. No Ombuds complaint file or case record was obtained, so every disposition above remains the office's self-reported classification. And some FY25 chart values were read from page screenshots rather than extracted text, so a residual transcription risk attaches to each figure taken from that chart. Whether the Board has ever asked for additional verification of the annual report's statistics is not established by any reviewed document.