Summary

  • The Address Supporting Organization's Address Council is described across ASO and ICANN pages as the body that manages and conducts the selection of ICANN Board Seats 9 and 10, while ICANN publishes the dated announcements marking each visible stage.
  • The two dates on offer measure different things. 18 September 2024 is identified as an appointment to Seat 10; 3 June 2025 is identified as a nominations notice for Seat 9 in the 2024-2025 cycle. Neither is a nomination window, a recorded vote, or a term boundary.
  • ASO-appointed terms are described as generally or typically three years, which yields no computable expiry without a commencement rule and the event that starts the clock.

Two dates, two different kinds of news

On 18 September 2024, ICANN's announcement feed carried an item identified as the Address Council appointing Tom Slater to ICANN Board Seat 10 (ICANN announcement, 18 September 2024). Nine months later, on 3 June 2025, it carried a different kind of item: a notice of Address Council nominations for Seat 9, labelled for the 2024-2025 cycle (ICANN announcement, 3 June 2025).

Read quickly, those two lines look like an institution keeping its house in order: a seat filled, a seat opened. Read slowly, they are two instruments of different types, and the difference matters more than the resemblance. One is an endpoint reported after the fact. The other is an opening reported before it. A rotation that publishes endpoints and openings separately, without a page that joins them, asks readers to reconstruct the middle.

What the retained material cannot tell you is how much middle there is. The summary retained for the September 2024 announcement does not carry the nomination window that preceded the appointment, the deciding body's vote or tally, the effective appointment date, or the end date of the term. The announcement text itself was not obtained in this review, so I cannot tell you whether those fields are absent from the announcement or absent only from what I could see. That distinction is not pedantry. It is the whole question.

Endpoint and opening are not the same instrument

Consider what each date can legitimately support. A publication date on an appointment announcement can support the proposition that an appointment was announced on that date. It cannot support a claim about when the appointment took effect, because an announcement is not a commencement instrument. A publication date on a nominations notice can support the proposition that nominations were invited around that date. It cannot support a claim about when the window closed, because a notice is not a window.

This is where most secondary accounts of reserved Board seats go wrong. They compress a sequence of instruments into a single date and then present the date as if it were the event. The compression is invisible once it is written down: "appointed in September 2024" reads as a fact about a person and a seat, when in truth it is a fact about a web page.

A smaller illustration sits in the same material. A summary attached to the September 2024 announcement places Ron da Silva in Seat 9 from July 2022. No announcement text obtained in this review confirms that placement, and no term boundary for that tenure was obtained either. The sentence functions as a helpful orienting note and as a warning: chronology assembled from adjacent summaries looks like a schedule and is not one.

Who actually holds the pen

The ASO's own pages are unambiguous about the process owner. The ASO is described as responsible for selecting individuals to serve in ICANN Board Seats 9 and 10, with the Address Council managing the selection and appointment process (ASO board appointments). A dedicated procedure page describes the Council managing the appointment process under an established selection procedure (ASO selection procedure), and a seat-specific page describes the ASO managing the selection for Seats 9 and 10 (ASO Board Seats 9 and 10). The ASO site itself is identified as the publication point for the Council's Board-seat materials (aso.icann.org).

The Number Resource Organization appears in a supporting role, described as coordinating documentation of the ASO's role and describing ASO-appointed seats as typically served on three-year terms (NRO: ASO ICANN Board seats). ICANN, meanwhile, is the publisher of the two dated announcements. Those roles are not the same role, and the fact that both institutions are credible does not make them interchangeable.

The governing instrument is identified as the ICANN Bylaws, which set Board composition, the ASO-associated seats and director terms (ICANN Bylaws). The Bylaws text was not obtained in this review, so no article, clause or provision wording is verified here. A Board roster page is identified as the natural secondary cross-check on who currently occupies an ASO-associated seat (ICANN Board members); that page was not obtained either. So the two documents that would convert institutional description into verified structure — the Bylaws and the roster — are both named and both outside this review.

A reader is entitled to ask why that matters when the ASO's own pages agree with each other. It matters because agreement among pages published by the process owner establishes what the process owner says about itself. It does not establish the term a director actually serves, the date a term begins, or the mechanism by which one Council decision becomes one Board seat. Those are the questions a lawyer, a competing candidate, or a registry planning five years of continuity actually has.

What a three-year term is not

Two sources describe the ASO-appointed term as generally or typically three years (ICANN board seats 9 and 10; NRO: ASO ICANN Board seats). Take the description at face value and notice how little follows from it.

A duration is not a date. To turn three years into an expiry you need a commencement rule — does the term run from the appointment, from the appointment's effective date, from the first Board meeting after it, or from the end of the predecessor's term? — plus the event that triggers it. Neither the rule nor the event was obtained in this review. So the honest statement is not that Seat 10 expires in 2027.

It is that a reader holding only the published dates cannot compute when Seat 10 expires at all, and anyone who writes "three years from September 2024" is performing arithmetic on an announcement rather than reading a record.

The same discipline applies to the two seats together. It is common in reserved-seat arrangements for terms to be staggered so that one seat turns while the other holds. Whether these two seats are staggered, and on what cycle, was not established in the reviewed material. The two dates in hand do not settle it: an appointment in September 2024 and a nominations notice in June 2025 are consistent with several arrangements, including arrangements that are not staggered at all.

The chain a reader would need

To audit a rotation you do not need a single grand register. You need six links, and you need them to point at each other. First, the applicable version of the selection procedure, with its effective date. Second, the nomination opening and closing dates. Third, a decision record naming the deciding body, its method and the date. Fourth, the appointment notice. Fifth, the effective appointment date. Sixth, the term start and end.

Any one of those links can live on a different page. That is fine. What defeats audit is not distribution but disconnection: an endpoint announcement with no pointer to the procedure it applied, a nominations notice with no successor page, a term described as typical with no commencement rule anywhere in the chain. In the material reviewed here, none of the six links was obtained in a form that could be read and verified. That is a statement about this review's reach, not about the institutions' filing cabinets.

It is worth being precise about what the absence of those links does and does not prove. It does not prove that a required step was skipped. It does not prove that any record was withheld or obscured. It does not prove that either selection was invalid, and it does not prove that the Address Council's authority is defective. What it does prove is narrower and still consequential: an outsider cannot reproduce the lifecycle of either seat from the published material, and must either trust the summary or say plainly that the chronology is unfinished.

Where responsibility sits

Responsibility should follow control, and control here is divided unevenly. The Address Council holds the process record: the procedure, the nominations, the decision. It is therefore first in line for the traceability question, because it is the body that could publish the chain without asking anyone's permission. ICANN holds a narrower but real position: it publishes the announcements, and it chooses whether a notice links backward to a procedure and forward to an outcome. That is a publication duty about clarity and linkage, not a claim that ICANN supervises the Council's selection.

Readers come last, and should be assigned the least. A reader can be asked not to convert a publication date into a process date. A reader cannot reasonably be asked to reverse-engineer facts that only the process administrator holds. When commentary implies that the public should simply "follow the process more closely," it is quietly transferring an institutional obligation onto an audience that has no access to the file.

Two limits on this reading belong beside it. First, the phrase "community administration" is not self-validating. A label describing who administers a process does not by itself establish the mandate, the review path, the conflict rules or the appeal mechanism; those must be found in instruments and read. Second, an incomplete public chronology creates an informational advantage for the body that holds the records even where nobody intended concealment. That is a structural observation about incentives, not a finding of misconduct, and it should be stated as the former.

The consequence for anyone relying on the seats

Put yourself in the position of an operator, a candidate, or a registry whose continuity planning runs through ICANN's Board. You want to know two things: who holds Seat 9 and Seat 10 now, and when each turns. The reviewed material answers neither with a date you can check. It gives you an appointment announcement, a nominations notice, and a duration described as typical.

That is enough to write a news brief and not enough to plan against. A candidate deciding whether to prepare for the next cycle cannot tell from the published dates whether the window opens this year or next. An operator mapping Board-level risk cannot tell whether both seats turn in the same year. A journalist cannot tell whether the June 2025 nominations notice for Seat 9 closed with an appointment, because no outcome for that cycle was obtained in this review.

The record itself has a home in our directory, where the entry for the selection sits alongside the institutions that administer it: ICANN Board Seat 9 and 10 Selection Only for Strategic Circle.

What would have to change

Three developments would change what this record means, and each is small enough to be published in an afternoon.

A nominations page that states an opening date and a closing date, and links to the procedure version it applies. An appointment notice that states an effective date and names the term it fills. A term schedule, even a plain table, that states for each seat when the current term began and when it ends.

If those existed, this article would be shorter and duller: a description of a routine rotation, competently documented. That is the point. Dullness is what an auditable rotation looks like from outside.

If the missing records instead surface elsewhere — in Council minutes, in a Bylaws provision I could not obtain, in a roster that carries term dates — then the correct response is to correct the boundary drawn here, not to defend it. A record boundary is a claim about what was reviewed. It should be as easy to revise as it was to state.

Until then, the honest formulation is the uncomfortable one. Two ICANN Board seats reserved for the address community turn on announcements. Whether they turn on schedule is a question the announcements were never designed to answer.