Summary

  • The Address Supporting Organization publishes its own account of what it and its Address Council do: it develops recommendations on Internet Protocol address policy, advises the ICANN Board, and defines procedures for selecting individuals to serve on ICANN bodies. Each of those duties runs toward ICANN. None of them, as published, runs toward a regional registry's conduct.
  • The ASO's constitutional home is an article of ICANN's own bylaws, which is to say its existence is a construct of the parent organisation rather than of any instrument the registries concluded among themselves.
  • The documents that would settle whether a recognition, review or derecognition power exists over a registry are named and identifiable — a 2004 ICANN-hosted memorandum, an organisational-review index, a board-materials index — and none of them returned readable text in this run. The silence in this article is a fact about retrieval on 17 October 2026, not a finding that the power is absent.

There is a particular kind of gap that appears when an institution describes itself well and describes its limits not at all. The description is accurate, quotable, and short. The reader who needs to know what happens when a registry behaves badly finds the description complete on its own terms and silent on the question actually asked. That is the shape of the record examined here.

The Address Supporting Organization is, in its own words, "one of ICANN's three Supporting Organizations (SOs)". BTW's directory entry for the Address Supporting Organization Address Council records the object's topic as internet number-resource governance and RIR recognition policy — a topic that promises a supervisory relationship. This article went looking for the text underneath that promise.

What the record actually contains

The retrieval record for this article is short enough to state in full. Two first-party fragments were obtained. The first is a provider-reported excerpt from the ASO's own portal page, and its ellipses are the provider's, not an editor's:

The ASO is one of ICANN's three Supporting Organizations (SOs). We review and develop recommendations on Internet Protocol (IP) address policy and advise the ICANN Board on policy issues... The ASO AC... Defining procedures for the selection of individuals to serve on ICANN bodies.

The second is a heading from the ICANN Bylaws:

ARTICLE 9 ADDRESS SUPPORTING ORGANIZATION

That is the whole of it. Two fragments, one of them four words of capitalised headings. Everything else this article says about what those documents do not establish has to be read against that thin base, which is why the boundary section at the end of this piece is not a formality.

The sentence that defines the mandate

The ASO excerpt is worth reading slowly, because three distinct functions are packed into it and the direction of each is the entire subject.

First, "We review and develop recommendations on Internet Protocol (IP) address policy." The object of the verb is policy, not conduct. A body that develops policy recommendations is doing legislative work of a consultative kind; it is producing the rules that others will apply.

Second, "advise the ICANN Board on policy issues." The recipient of the advice is named and it is the Board. Advice is a real power — it shapes what the deciding body decides — but it is a power exercised before a decision and by someone who is not the decider. It creates influence, not authority to command.

Third, "Defining procedures for the selection of individuals to serve on ICANN bodies." Here the Council writes process: it sets out how people are chosen for seats elsewhere in the ICANN structure. It is a rule-writer for appointments into a different institution's organs.

Read together, the three functions form a consistent picture of a coordination body oriented toward the top of the structure. Policy flows up. Advice flows up. Personnel procedures flow up into ICANN's own bodies. Nothing in the published list flows outward and downward toward the entities whose conduct a reader worried about registry accountability would want examined.

This is a reading, and it should be labelled as one. An enumerated list of functions is evidence of a document's emphasis, not a complete inventory of a body's legal powers. Institutions routinely hold powers that their public-facing pages do not bother to describe, and a homepage is the weakest possible evidence of the absence of an authority. What the reading establishes is narrower and still useful: if a supervisory or derecognition power exists, it is not among the things the Council chooses to tell the public it does.

Where the ASO lives: inside ICANN's own bylaws

The second fragment is a heading, and headings are usually the least interesting part of a legal document. This one carries a structural fact. The Address Supporting Organization has a dedicated article in the ICANN Bylaws — Article 9 — titled for it by name.

That placement is not decorative. It means the ASO's constitution, its reason for existing in the shape it exists, is a provision of the parent organisation's constitutive document rather than a treaty among the recognised parties or a free-standing corporate charter. The supporting organisation is an organ of ICANN, created by ICANN's own rules, and amendable through ICANN's own processes.

The consequence for the accountability question is direct. If the ASO's powers are conferred by an article of the ICANN Bylaws, then the question of whether the Council can act against a registry is, at least in part, a question about ICANN's own constitutional text — and that text is precisely what this run did not read beyond the heading. The Bylaws are a public document and the article is a public article; the failure here is one of retrieval, not of secrecy.

The distinction matters for how a reader reasons about remedies. Authority granted by a parent organisation's bylaws can be amended by that organisation through its own amendment procedure. Authority granted by contract among the recognised parties requires those parties' consent to change. Those are different political economies. The heading alone tells us which instrument contains the ASO; it does not tell us which of the two political economies governs a recognition decision, because the answering text is in the article body, and possibly also in the memorandum.

The memorandum that would settle it

ICANN hosts a page titled Address Supporting Organization (ASO) Memorandum of Understanding with a reported publication date of 21 October 2004. The date is a metadata value attached to the page, not something this article verified against the document's own text, because the document's text was never returned.

A provider summary accompanying the ASO portal excerpt describes the ASO as operating "under an MoU between ICANN and the Number Resource Organization (NRO)". That sentence is a description, not a quotation from the memorandum, and it should be read as such. It is consistent with the separately retrieved page title, since a memorandum between ICANN and the Number Resource Organization concerning the Address Supporting Organization is what that title describes. But consistency between a summary and a title is not evidence about content, and this article will not treat it as such.

What a memorandum of understanding between a corporation and an unincorporated association of registries would ordinarily contain is a recital of what each party recognises, what each undertakes to do, how the arrangement is reviewed, and how it ends. Whether this particular memorandum contains those elements, and in what language, and with what thresholds, is unknown here. So is whether it has been amended, superseded, or replaced by a later text. A document dated 2004 sitting on a 2026 website may be the operative instrument or may be an archive copy retained for the record; the retrieval gave no way to tell.

This is the point at which a less careful article would assert the answer. The structure invites it: a memorandum between ICANN and the registries' own association sounds like the place where recognition would be defined, and a reader primed by the directory topic would accept the inference. But an inference about the contents of an unread document is not reporting, and the practical harm of getting it wrong is real — counterparties make decisions about registry risk on the basis of what they believe the instruments provide.

Why direction matters more than existence

Suppose, for the sake of testing the reasoning rather than asserting a fact, that a recognition or derecognition provision does exist somewhere in the unread material. The published description would still be telling us something important: that the Council's self-presentation is oriented toward policy and personnel rather than toward enforcement.

Direction is the useful variable because it governs what a reader can rely on without reading further. A mandate that runs upward — advice to a board, recommendations on policy, procedures for selecting officials — can be discharged in public, in meetings, in published recommendations, and its exercise is largely visible. A mandate that runs outward at a registry's conduct would have to be discharged through findings, deadlines and consequences, and its exercise would leave a decision record.

The absence of any such function from the published list is therefore not proof that the enforcement track does not exist; it is evidence that the enforcement track, if it exists, is not part of what the Council presents as its work.

That distinction has a practical use even under uncertainty. A counterparty assessing exposure to a registry's behaviour should not treat the existence of a global coordinating body as an assurance that misconduct will be corrected. Coordination bodies coordinate. Whether they also discipline is a separate question that requires a separate instrument, and the instrument is the thing to demand.

The composition question that cannot yet be judged

The concern that motivated this commission was conflict of interest: a body whose members are appointed by the same entities whose recognition it might adjudicate cannot credibly adjudicate them unless the operative text supplies something extra — an independent appointment route, a disqualification rule, a supermajority, or an external check.

The second retrieval attempt returned a provider characterisation that the Address Council consists of members appointed through regional processes. That is a summary sentence, not quoted text, and it does not identify the appointing instruments, the terms, any recall right, or any voting threshold. So the conflict question cannot be answered here in either direction. It cannot be shown that the conflict exists in the text, and it cannot be shown that the text guards against it.

What can be said is that the conflict is a textual question rather than a moral one. Every body that judges its own constituents faces it. The difference between a body that manages the conflict credibly and one that does not is whether the founding instrument allocates the decision in a way that survives the alignment of interests: who may file, who screens, who votes, what majority, and what happens when the vote fails. None of those five questions has an answer in the material this run holds, and each of them is answerable only from the instrument.

What a counterparty should ask instead

The practical output of a negative finding is a better question. Three are worth carrying into any assessment of registry risk.

First, ask which instrument confers recognition and what it says about withdrawal. Not whether a registry is recognised — that is observable from the registry's operation — but which document makes recognition a decision rather than a fact, and what procedure that document sets out for reversing it.

Second, ask who the decider is and by what majority. An answer of the form "the community" or "the supporting organisation" is not an answer; it is a placeholder for a procedure that either exists in writing or does not exist.

Third, ask what public record a decision would leave. A supervisory power exercised without a published finding is indistinguishable, from outside, from no power at all, and the reader's inability to distinguish them is itself a governance fact worth naming.

These questions have the advantage of being answerable by document production. They do not require anyone to characterise an institution's culture, and they do not depend on the goodwill of the party being assessed.

Evidence boundary, 17 October 2026

Three retrieval attempts were made against the operative instrument class on 17 October 2026, and the record of them belongs in the article rather than in a footnote.

The first attempt returned no page text at all: the primary provider reported capacity unavailable after a bounded wait, and the fallback returned five candidate addresses whose excerpts were empty, each marked as identified but not retrieved. The second attempt succeeded as a request and returned only the two fragments quoted above; the memorandum and the organisational-review index, both supplied explicitly in that request, produced no page text. The third attempt targeted a different document class — the registries' side of the Council appointment route — and returned an empty candidate set.

A standing rule ends retrieval after two consecutive failures on one instrument class, so the memorandum, the organizational-review record and the board-materials index were not pursued further. None of the three is characterised in this article. Each is named so that a reader can go to it directly, which is the only honest use of a document one has not read.

The eight propositions this article explicitly does not assert are worth stating once, compactly. It does not assert that the Council lacks any power over registry recognition or derecognition, nor that it has never exercised one. It does not assert that the 2004 memorandum is current, unamended, or the only instrument constituting the ASO. It does not assert that Article 9 contains no provision bearing on recognition or composition. It does not assert that an organisational review occurred, what it recommended, or what the Board did with it. It does not assert the size, term or appointment route of the Council.

It does not assert that no procedure has ever been opened against a registry. And it does not assert, beyond the ASO's own sentence, that three is the present number of ICANN supporting organizations.

What it does assert is narrow, sourced and checkable. The ASO describes itself as one of ICANN's three supporting organizations, describes its work as policy recommendations and advice to the Board, and describes the Council as defining procedures for selecting individuals to serve on ICANN bodies. A dedicated article of the ICANN Bylaws is titled for the ASO. ICANN hosts a page titled as a memorandum of understanding for the ASO with a reported 2004 publication date. And on the one description of the Council's duties that this run could read, no duty points at a registry.

That is a small finding. It is also the finding that survives contact with the record, and the difference between the two is the whole discipline.

Sources