The ASO MoU's Two-Level Representation Problem

Summary

The ASO MoU creates a two-level representation chain: the RIR system is represented through the NRO, and the ASO Address Council consists of the NRO Number Council. Each region supplies three council members, two through regional community selection and one through the RIR board.
That chain is workable for defined ASO tasks, but it does not prove broad global operator consent. Low participation, professionalized meeting culture, incumbent networks, board appointments and regional membership filters can be carried upward into ICANN with a stronger institutional label than the underlying participation supports.
The solution is not to abandon the ASO. It is to bound its scope: process review, global policy transmission, new-RIR recommendations, ICANN Board selection and advice should each carry clear participation evidence, conflict disclosure, regional denominator reporting and limits against claiming more authority than the selection chain can justify.

Representation at two removes

The Address Supporting Organization is one of the quietest high-consequence structures in internet governance. It rarely produces the public drama associated with domain-name controversies, but it sits on the route by which the number-resource community participates in ICANN. It advises the ICANN Board on internet number-resource policy. It helps handle global policy development. It provides recommendations about recognition of new regional internet registries. It selects people for ICANN Board seats. It gives the RIR system an institutional place inside ICANN.

The problem is not that the ASO exists. The problem is how easily its name can make representation appear more direct than it is. The ASO is not a global electorate of network operators. It is a structure built through the RIRs and the Number Resource Organization. Under the 2004 ICANN Address Supporting Organization MoU, the NRO fulfills the ASO role. The ASO Address Council consists of the members of the NRO Number Council. The ASO Address Council page repeats that the council consists of fifteen NRO Number Council members and that each RIR community appoints two members while each RIR executive board appoints one.

That creates representation at two removes. First, operators and other affected parties participate through regional RIR communities, membership structures, policy forums and elections. Second, those RIR-selected people sit as the NRO Number Council and therefore as the ASO Address Council. The result is not illegitimate by default, but it is indirect. The chain deserves scrutiny every time the ASO is treated as the voice of the numbering community.

Indirect representation can be appropriate when the mandate is narrow. A global address policy cannot be developed by a single mass vote of everyone who uses the internet. The number-resource system is technical, regionally administered, legally plural and operationally dependent on RIRs. The RIR communities are the places where policy expertise, registry experience and affected operators are most likely to meet. A council drawn from those communities can provide continuity and knowledge.

But indirect representation becomes risky when the mandate is rhetorically expanded. A small group selected through regional mechanisms may be well suited to verify whether each region followed its policy procedure. It may be less suited to claim broad public consent on contested economic, political or institutional questions. A council whose members are selected through RIR-specific processes may be able to choose ICANN directors under a defined procedure. It should be cautious about suggesting that those directors were chosen by a global operator electorate.

The ASO MoU itself points toward boundedness. It defines roles and processes supporting global policy development, mechanisms for recommendations on recognition of new RIRs, and procedures for selecting individuals to serve on ICANN bodies. It does not create a general legislature for number-resource users. It does not define a global operator franchise. It does not say that the Address Council can decide regional registry disputes or impose ordinary operator obligations.

The two-level problem is therefore a legitimacy problem rather than a formal invalidity problem. The ASO can be legitimate for its defined tasks and still overstate itself if it treats RIR-selected representation as complete global consent. The more concentrated or low-participation the underlying regional processes are, the more important this distinction becomes.

Fifteen seats do not solve the denominator problem

The ASO/NRO Number Council has a clean arithmetic design. Five regions. Three seats per region. Two selected by the regional community. One appointed by the RIR board. Fifteen volunteers. Current ASO and NRO pages present the structure clearly, with regional terms and appointment markers. Equal regional weight is easy to understand and easy to administer.

But a clean seat count does not solve the denominator problem. Representation always depends on who had a meaningful chance to participate, who actually participated, and what interests they carried. Three seats from a region do not automatically represent every operator, resource holder, public agency, enterprise user, civil society voice, small network, hosting provider, legacy holder or customer in that region. The seats represent the output of that region's selection mechanism.

The denominator problem is especially sharp in number-resource governance because the affected universe is larger than the active community. Operators depend on registry services even if they never attend RIR meetings. Enterprises use assigned space even if they do not follow policy lists. Cloud customers depend on resource-holder decisions without knowing the registry layer. Governments depend on registry data for security and continuity, but they may participate only intermittently. Small networks may lack time, travel budget, policy expertise or staff capacity. Legacy holders may have legal interests but limited institutional involvement.

None of these groups disappears merely because a regional election or appointment took place.

The IETF's RFC 7020 describes the Internet Numbers Registry System as a hierarchy rooted in IANA, serving RIRs, which then serve local registries and other customers, and it identifies uniqueness and registration accuracy as central operational goals. That hierarchy explains why representation is hard. The affected parties are not all members of the same institution. Some are customers of customers. Some are downstream users. Some are affected through routing, abuse response, financing or market reliance. A regional policy forum may be open, but openness is not the same as broad attendance.

The ASO structure handles this by relying on RIR communities as the most practical selection base. That may be reasonable, but it should be named as a practical compromise. When the Address Council acts, the public record should not imply that fifteen seats have measured the full affected population. It should say which regional selection routes were used, what turnout or participation evidence exists where available, what appointment choices were made by boards, and what consultation occurred for the specific matter.

Equal regional weight also hides unequal population, network, economic and membership denominators. AFRINIC, APNIC, ARIN, LACNIC and RIPE NCC regions differ in population, number of networks, resource holdings, member counts, market concentration, state involvement, IPv4 scarcity exposure, transfer-market activity and meeting culture. Three seats per region can protect smaller regions from being overwhelmed by larger ones. It can also overrepresent or underrepresent affected operators depending on the question. A global policy about IANA allocation may justify equal regional process weight.

A question about operator impact, board selection or market reliance may need additional evidence.

The problem is not solved by saying that the council members are volunteers or experts. Expertise is valuable, but expertise is not a denominator. Nor is the problem solved by saying that procedures are open. Open procedures still require awareness, time, language access, travel or remote participation, trust and perceived relevance. Low participation can be perfectly compatible with formal openness.

This is why ASO legitimacy should be measured function by function. For global policy review, the key evidence is whether all regional policy processes were followed and significant viewpoints were considered. For ICANN Board selection, the key evidence is candidate outreach, conflict disclosure, comment participation, selection criteria and voting record inside the council. For new-RIR recognition advice, the key evidence is regional support, service continuity, affected network backing and ICP-2 criteria. One fifteen-seat structure cannot answer all those questions by itself.

The regional community seat and the board-appointed seat are different mandates

The ASO/NRO Number Council combines two kinds of selection. Each RIR community selects two people, and each RIR executive board appoints one. This blend can be defended. Community-selected seats help connect the council to open regional participation. Board-appointed seats can add continuity, institutional knowledge or skills that elections might miss. The problem comes when the two mandates are treated as if they are identical.

A community-selected seat and a board-appointed seat carry different legitimacy claims. The community-selected member can claim a route through the region's open selection procedure, though the strength of that claim depends on participation. The board-appointed member can claim institutional appointment by the RIR's governing board, though the strength of that claim depends on the board's own accountability, criteria and transparency. Neither claim is invalid. They are simply different.

The distinction should matter in public reporting. When the Address Council reaches a conclusion, readers should know whether a regional view was carried by elected members, board appointees, or both. When a conflict arises between community sentiment and board preference, the difference becomes even more important. A board appointee may be well qualified, but a board appointment is not a substitute for community selection. A community-selected member may have stronger electoral legitimacy, but a low-participation election may not express a broad regional consensus.

The NRO page on the ASO AC/NRO NC marks appointed members and states that each RIR community selects two people while each RIR executive board appoints one. The ARIN elections page gives a concrete example of regional mechanics: for two out of three years, eligible ARIN General Members elect one ARIN representative to the NRO Number Council; every third year, the ARIN Board appoints one from a community-developed pool of nominees. It also says each ARIN member organization is entitled to designate one voting contact per organization and must be a General Member in Good Standing by the eligibility date to participate.

That example shows why denominator disclosure is essential. The ARIN route is not "all North American operators voted." It is a defined membership route through eligible General Members, voting contacts and a calendar. That may be the correct route for ARIN's structure, but it should not be converted into broader language than it supports. Other RIRs have their own selection rules. The ASO's public legitimacy depends on making those differences legible rather than smoothing them into a generic "community" label.

Board-appointed seats also raise a concentration risk. RIR boards are themselves selected through regional institutional mechanisms that may favor active members, incumbents, professional entities or well-resourced networks. When a board appoints one of three council seats, it can improve competence, but it can also reinforce the same networks that already dominate participation. That risk is not proof of capture. It is a reason to require criteria, open nomination, conflict disclosure and public reasoning.

The risk is magnified because the Address Council does not merely discuss policy. It also defines procedures to select individuals for ICANN bodies, including ICANN Board seats, and selects ICANN directors for ASO seats. A board-appointed member of the NRO Number Council may therefore participate in choosing ICANN directors. That is a legitimate function under the MoU, but it makes the chain of selection worth examining. A board appointment inside an RIR can influence a board appointment inside ICANN.

The answer is not to abolish the appointed seat. The answer is to name its mandate honestly. A board-appointed member should be described as adding institutional appointment, not as a direct community electoral outcome. The public record should show who appointed the member, what criteria were used, what conflicts were disclosed, how the community could comment, and how the member participates in votes. That would preserve the value of appointment while reducing the risk of hidden amplification.

Low participation is amplified, not averaged away

The central representational flaw is amplification. If a regional process has low participation, the ASO structure does not average that problem away. It carries the result upward. A candidate selected by a narrow regional electorate becomes one of fifteen members of a global-facing council. A view formed in a small policy forum can become part of an ASO process report. A board-appointed member chosen from a small professional circle can vote in a selection process for ICANN Board seats. The institutional label becomes larger than the participation base.

This matters because internet-number governance is a specialist domain. The people who attend RIR meetings, read policy lists, nominate candidates, understand global policy definitions and follow ASO minutes are often a small subset of those affected. They may be highly competent and public-spirited. They may also share assumptions, professional networks, travel patterns, employer types or market interests. Concentration can be social rather than conspiratorial. A room can be open and still be narrow.

Low participation changes the meaning of consensus. In a small active community, silence can be interpreted as acceptance. But silence may mean that affected parties did not know, did not understand, could not attend, lacked translation, saw no practical route to influence, or assumed the issue was too technical. The ASO structure should not treat every unopposed regional outcome as broad consent unless participation evidence supports that reading.

The global policy development procedure tries to address part of this problem by requiring regional consideration and Address Council review of whether significant viewpoints were adequately considered. The 2004 ASO MoU's attachment gives the Address Council a process-review role before forwarding global policy to ICANN. The ASO operating procedures similarly describe review of whether each region's policy development process was followed and whether there were process failures. That is useful, but process review is not the same as participation depth.

A region can follow its formal process and still have a thin record. It can post to a mailing list and receive little response. It can hold an open session attended mainly by recurring entities. It can complete an election with a small electorate relative to affected networks. It can document minutes without documenting who was absent. The ASO should therefore separate process compliance from participation quality. Both matter.

The same is true for ICANN Board selection. The Address Council may follow its selection procedure, run nominations, interviews, comment periods and votes. That proves procedural compliance. It does not automatically prove that the global operator community meaningfully influenced the choice. Board seats are high-value because they affect ICANN's mission, budget, accountability, IANA-related oversight and cross-community decisions. A selection route that begins in RIR-selected council membership should be especially careful about recording outreach and participation.

The practical fix is a participation statement for every major ASO act. For a global policy recommendation, the statement should list regional discussion venues, dates, participation measures where available, unresolved objections, staff involvement, community comments, and the Address Council's process findings. For a Board-seat selection, it should list candidate outreach, application numbers, public-comment activity, conflicts, abstentions and the council vote. For a new-RIR recognition recommendation, it should list affected-region support, service-area evidence, member backing, operational readiness and dissent.

This is not bureaucracy for its own sake. It is a way of preventing the ASO label from laundering thin participation. If participation is broad, the statement proves it. If participation is thin, the statement helps future work improve. If interests are concentrated, disclosure allows readers to judge whether the council stayed within its mandate. If a decision must proceed despite thin participation because operational continuity requires it, the record can say why.

Amplification is unavoidable in any representative system. The issue is whether amplification is visible. The ASO can remain useful if it treats participation as evidence, not as a slogan.

Concentrated interests can look like continuity

Number-resource governance rewards continuity. The people who understand the system often have long histories in RIR meetings, operator groups, policy lists, technical standards, transfer markets, routing security, registry operations or ICANN structures. That institutional memory is valuable. It can prevent naive policy design. It can help distinguish real operational constraints from political demands. It can keep global numbering work from being captured by governments or short-term market actors.

But continuity can also protect concentrated interests. The same professionals who know the system best may work for large operators, registry service providers, consultancies, brokers, vendors, research networks or institutions with repeated access. They may not be corrupt. They may simply be present when others are not. Over time, presence becomes reputation, reputation becomes nomination, nomination becomes selection, and selection becomes authority.

The ASO/NRO Number Council is vulnerable to this dynamic because it relies on regional participation pools. If the active pool in a region is dominated by a particular kind of actor, the council may inherit that bias. Equal regional seats do not fix that if each region has its own concentration pattern. One region may be dominated by large access providers. Another may be dominated by national internet registry structures. Another may be shaped by transfer-market entities. Another may have strong staff influence because community participation is thin. Another may have high technical expertise but weak small-network representation.

This matters most where the ASO's work touches ICANN Board selection and global policy. The Address Council's role in selecting ICANN directors gives it influence beyond the narrow policy-review function. An ICANN director selected through the ASO route serves on the ICANN Board, not as a delegate with instructions from the council. The director's decisions affect ICANN as a whole. If the selection pool is narrow, concentrated interests can gain durable influence in ICANN's central governance without ever winning a broad global vote.

The same risk appears in global policy timing. A policy that appears technical may have distributional consequences. Rules for allocation after IPv4 exhaustion, ASN blocks, returned address space, registry accuracy, RPKI-related services or future registry-system evolution can affect market value, operational cost and institutional leverage. The ASO's role may be procedural, but procedural judgments can still shape outcomes: whether significant viewpoints were considered, whether concerns require further review, whether a proposal is forwarded, and how ICANN is told to understand the record.

The answer is not to exclude experienced entities. That would damage the system. The answer is to disclose interests and measure representation. Council members should have clear conflict-of-interest declarations. Selection processes should identify employer and affiliation categories without turning them into quotas. Public reports should distinguish comments from large resource holders, small networks, governments, civil society, technical operators, commercial intermediaries and individual experts where feasible. Meeting records should show whether the same voices are repeatedly decisive.

The system also needs renewal mechanisms. Term limits, open nomination outreach, remote participation, multilingual materials, travel support, fellowships, small-network briefings, regional listening sessions and post-decision review can reduce concentration. These measures are not symbolic. They change who can show up and who can understand the stakes before a decision is effectively settled.

Concentrated interests become most dangerous when they are presented as neutral continuity. A closed circle can say it is protecting stability. Sometimes it is. Sometimes it is protecting its own familiarity, access or market position. The ASO should not be forced to prove bad faith before improving transparency. The better standard is structural: if a selection route can amplify concentrated interests, the public record should expose enough data to evaluate that risk.

Scope discipline is the main safeguard

The best protection against the two-level representation problem is scope discipline. The ASO should do what the MoU, bylaws and operating procedures authorize, and it should avoid speaking as if that authority extends to all number-resource questions. A narrow mandate can survive indirect selection. A broad mandate cannot.

The first scope is global policy. The ASO is involved in global policies that require agreement of all RIRs and specific action by IANA or another ICANN-related body. That definition is narrow. It should not be expanded to ordinary regional policy or to any issue that happens to interest multiple RIRs. If a policy does not require a global IANA or ICANN-related action and common RIR agreement, it should remain outside the global policy route.

The second scope is process review. The Address Council's global policy role is largely to check whether the regional procedures were followed and whether significant viewpoints were adequately considered. It should be cautious about substituting its own policy preferences for regional outcomes. If it has concerns, it can refer the matter back for regional consideration. That is different from becoming a superior policy chamber.

The third scope is new-RIR recognition advice. The ASO can provide recommendations to the ICANN Board concerning recognition of new RIRs according to agreed requirements and policies such as ICP-2 or its successor. That task is high consequence, but it is still criteria-based. It should not become a way for current RIR insiders to block competition or preserve institutional territory unless the criteria justify the recommendation. Recognition advice should include affected-region support, operational readiness, service continuity and non-discrimination evidence.

The fourth scope is ICANN body selection. The Address Council can define procedures and select individuals for ICANN roles, particularly Board seats 9 and 10. This role is formal and important. It should be treated as a selection mandate, not a general power to instruct those directors after selection. Once selected, ICANN directors owe duties to ICANN. The ASO's legitimacy depends on choosing well, not on pretending to control the Board through them.

The fifth scope is advice. The ASO can advise the ICANN Board on number-resource allocation policy in conjunction with the RIRs. Advice is influential, but it is not binding law. It should identify whether it reflects a formal council vote, RIR consensus, a regional policy outcome, staff analysis or a consultation summary. Advice that affects operators should be transparent about its evidentiary basis.

Scope discipline also means avoiding title inflation. A phrase like "the numbering community believes" should be used only when the record supports it. Often the honest phrase is narrower: "the Address Council concluded," "the RIRs agreed," "the regional policy processes produced," "the council's process review found," or "the current active entities supported." These distinctions may seem small, but they prevent legitimacy from being borrowed without proof.

Scope discipline should be written into public templates. Every ASO statement could identify capacity, authority source, affected function, participation evidence, vote or consensus method, conflicts, limits and next review. Such a template would make it difficult to overclaim unintentionally. It would also help outsiders understand what kind of decision they are reading.

The ASO can remain influential if it is modest. Modesty does not mean silence. It means matching claim to mandate. The more indirect the selection chain, the more important that match becomes.

ICANN should not overread the ASO's representational signal

ICANN also has a responsibility. It should not treat the ASO as if it magically solves representation for all internet-number questions. ICANN's bylaws say the ASO advises the Board with respect to policy issues relating to the operation, assignment and management of internet addresses, and that the Address Council consists of NRO Number Council members. ICANN's mission says it coordinates allocation and assignment at the top-most level and facilitates global number registry policies by the affected community and other related tasks as agreed with the RIRs.

These clauses require ICANN to respect RIR-community processes, but they do not allow ICANN to ignore participation quality.

When ICANN receives ASO advice, it should ask what the advice represents. Is it a formal global policy proposal? Is it a process review? Is it a recommendation on new-RIR recognition? Is it a Board-seat selection result? Is it a letter from the NRO Executive Council? Is it a view of the Address Council? Is it a joint RIR position? Each deserves different weight.

ICANN should also ask whether the relevant affected community was actually engaged. The phrase "affected community" appears in ICANN's mission context, but the affected community for numbering issues is not self-evident. Sometimes it is the RIR policy community. Sometimes it is resource holders. Sometimes it includes local internet registries, governments, downstream networks, route-security operators or users of public registration data. ICANN should not use the ASO as a shortcut around that analysis.

This is especially important for mission limits. ICANN is not a general regulator of internet services, and its own bylaws state that it does not hold governmentally authorized regulatory authority. If ICANN receives a broad ASO or NRO statement, it should not treat that statement as permission to move beyond top-level coordination, global number policy facilitation or tasks agreed with the RIRs. An indirect representative signal cannot enlarge ICANN's mission.

The Board-seat selection function makes this more sensitive. ASO-selected directors become ICANN directors. Their presence on the Board should not be taken as continuing consent from every RIR community for every Board action. Nor should their selection be treated as proof that the global operator community has endorsed unrelated ICANN positions. The selection route has a defined purpose. It does not replace public comment, cross-community deliberation or mission review.

ICANN can improve this by asking for participation reports with ASO submissions. It can ask for regional-denominator information where available, dissent summaries, conflict disclosures and evidence of outreach. It can publish how it weighed the ASO's input. It can avoid language that turns ASO advice into generic community endorsement. It can keep global number policy separate from domain-name policy controversies and other ICANN questions.

This would not disrespect the ASO. It would protect it. Overreliance can create backlash. If ICANN treats ASO advice as a complete democratic signal, critics will attack the ASO's indirect selection chain. If ICANN treats ASO advice as expert, regionally grounded and procedurally bounded, the advice remains valuable without carrying impossible claims.

The same restraint should apply in crises. If a troubled registry, recognition question or regional dispute reaches ICANN, ASO or NRO input may be essential. But ICANN should distinguish technical continuity, recognition criteria, regional community support, domestic legal issues and operator rights. ASO representation does not collapse all those categories into one answer.

The measurement agenda

The two-level representation problem can be improved only if it is measured. The ASO, NRO and RIRs should publish enough information to let outsiders evaluate selection depth and concentration without exposing private voting choices or personal data. Measurement should be routine rather than crisis-driven.

The first metric is eligibility denominator. For each community-selected seat, the region should identify who was eligible to nominate, support and vote. Was the electorate limited to members in good standing, meeting attendees, registered entities, policy-list subscribers, or a broader community? Were organizations given one vote, weighted votes, individual votes or consensus calls? What deadline determined eligibility? This tells readers what "community selected" means in practice.

The second metric is participation. How many eligible entities or individuals voted, commented, nominated or attended relevant sessions? What share of the eligible base participated where a denominator exists? If a region uses consensus rather than ballot, how many substantive comments were received, how many speakers supported or opposed, and whether silence was treated as consent? These numbers need not be perfect to be useful.

The third metric is affiliation. Without turning representation into rigid quotas, public reports should describe the broad affiliation mix of nominees, selected members and commenters: network operators, registry staff, government, academia, civil society, vendors, consultants, transfer-market actors, security community, small networks, large networks and unaffiliated experts. The goal is not to rank identities. It is to see whether one professional circle dominates.

The fourth metric is incumbency. How many selected people have prior ASO/NRO Number Council service, RIR board service, staff roles, advisory roles, frequent meeting leadership or employer links to current institutions? Experience can be positive, but repeated circulation should be visible.

The fifth metric is conflict and recusal. Council members participating in ICANN Board selection, new-RIR recognition advice or global policy review should disclose relevant employer, client, board, financial and close institutional interests. Recusals should be recorded. The aim is not to shame expertise; it is to distinguish expertise from stake.

The sixth metric is dissent. Every major ASO conclusion should summarize unresolved objections and minority concerns. If the council finds that regional processes were followed despite dissent, it should say what the dissent was and why the process still satisfied the standard. If participation was thin, it should say how that fact affected confidence.

The seventh metric is follow-through. After a global policy, Board selection, recognition recommendation or advisory statement, the ASO should publish what happened next. Did ICANN accept, reject, ask questions or take no action? Did regional communities revisit concerns? Did promised outreach occur? Representation is not only selection; it is accountability after the act.

None of these metrics requires an overhaul of the ASO MoU. Most can be implemented as reporting practice. The result would be a clearer record: not a claim that the ASO perfectly represents everyone, but evidence of how the ASO's representational chain actually functioned in each case.

The measurement agenda would also help identify when different tools are needed. If an issue shows low participation across regions but high operator impact, the ASO could call for additional outreach before forwarding a conclusion. If a Board-seat selection draws candidates from a narrow professional pool, the next cycle could broaden outreach. If a global policy attracts comments mainly from one interest group, the Address Council could flag the participation limit even if the formal process was followed.

Legitimacy improves when the system learns from its own denominators. Without measurement, the ASO can only repeat that its procedures are open. With measurement, it can show how open procedures performed.

A bounded mandate can still be strong

Critics sometimes assume that any call for scope limits weakens internet governance institutions. The ASO shows the opposite. Its mandate is strongest when it is precise. It can review global policy process because the MoU gives it that role. It can recommend on new-RIR recognition because the MoU names that function. It can select ICANN directors because the bylaws and procedures provide that route. It can advise on number-resource policy in conjunction with the RIRs because that is its defined place. These powers do not need inflated representational claims.

The ASO would be weaker if it claimed to speak for every operator on every registry question. That claim would be impossible to prove and easy to attack. It would turn every low-turnout election, board appointment or quiet mailing-list discussion into a legitimacy defect. It would invite governments and courts to ask why a private council claims public authority. It would invite operators to reject ASO advice when they never participated. It would make ICANN's mission boundaries less credible.

A bounded mandate avoids that trap. It says: this is a council selected through RIR routes; it performs defined tasks; it records participation; it discloses limits; it does not claim more than the selection chain can support. That is enough for many important functions. The internet does not need every institution to be a parliament. It needs institutions to be honest about the kind of authority they hold.

The two-level representation problem is therefore manageable. The first level, regional selection, can be improved through clearer eligibility, participation reporting, outreach and conflict disclosure. The second level, ASO/NRO representation inside ICANN, can be improved through capacity labels, scope limits, function-specific evidence and ICANN's own discipline in weighing advice. Neither requires abandoning the RIR model. Both require refusing to let the word "community" do all the work.

There is also a positive reason to keep the ASO. Number-resource policy is not ordinary public policy. It depends on technical uniqueness, routing constraints, registry accuracy, operational continuity, regional knowledge and global coordination. A body that gathers regional registry experience can prevent ICANN from making ill-informed decisions about numbers. The ASO's expertise is valuable. The problem is when expertise is mistaken for full representation.

The better frame is expert representative coordination. The ASO is representative because it is selected through RIR routes. It is expert because the active pool understands number resources. It is coordinating because it connects regional processes to ICANN. Each word limits the others. Expert does not erase representation. Representation does not erase indirectness. Coordination does not erase regional autonomy.

If the ASO embraces that frame, it can strengthen trust. Operators can see where to participate and what participation means. RIRs can show that their regional processes are not black boxes. ICANN can receive numbering advice without overclaiming democratic consent. Board candidates can be selected through a transparent path. New-RIR recognition can be considered against criteria rather than institutional protection. Global policies can be checked for process without becoming council-made law.

The risk is complacency. The ASO has existed long enough that its structure can feel natural. But legitimacy is not permanent. IPv4 scarcity, RPKI reliance, registry data accuracy, transfer markets, sanctions pressure, court disputes and regional governance crises have raised the value of number-resource decisions. The more valuable the decisions, the more important the representation chain becomes. What was acceptable as a small technical-community bridge in 2004 may need stronger evidence in a scarcity economy.

The conclusion is not that the ASO MoU failed. It solved a real coordination problem. The conclusion is that the solution created its own representation problem, and that problem should be managed openly.

The public test for ASO/NRO Number Council authority

A practical public test can keep ASO authority in proportion. First, identify the function. Is the council reviewing a global policy, recommending on new-RIR recognition, selecting an ICANN director, advising the Board, or organizing its own procedures? If the function is unclear, the authority is unclear.

Second, identify the selection chain. Which members participated? Which were community-selected and which were board-appointed? What regional processes selected them? Were any seats vacant, interim or recently appointed? Were conflicts disclosed? The public does not need private ballots, but it does need the chain.

Third, identify participation evidence. For the underlying issue, what regional discussions occurred? Who was eligible? Who participated? What views were significant? What objections remained? Were small networks, downstream operators, public-sector users or resource holders outside the usual policy circle reached? If the answer is thin, the conclusion should say so.

Fourth, identify the mandate limit. Does the decision bind anyone? Is it advice? Is it a process finding? Is it a nomination? Is it a recommendation? Is it a statement of regional consensus? Is it merely a council view? The public label should match the legal effect.

Fifth, identify the remedy. If a region believes the council misunderstood its process, if a candidate challenges a selection procedure, if an affected party believes conflicts were mishandled, or if ICANN overreads ASO advice, where can the concern go? A representative structure without remedy becomes a closed loop.

This test would make ASO governance more legible. It would also make criticism fairer. Critics could point to specific participation gaps rather than dismissing the entire body. Defenders could point to specific evidence rather than relying on institutional status. ICANN could explain why it gave particular weight to ASO advice. RIRs could improve weak regional participation before the next cycle.

The ASO MoU's two-level representation problem will not disappear because the number-resource system is inherently layered. But layered authority can still be accountable. It can show its source. It can state its limits. It can measure its participation. It can disclose its concentration risks. It can refuse to convert a small active community into a claim of universal consent.

That refusal is the mark of a mature institution. The ASO does not need to be everything. It needs to be exactly what its documents created: a bridge between RIR-based number-resource governance and ICANN, strong enough to coordinate defined tasks, modest enough to admit that the bridge is not the whole public.