Summary

  • The five RIR communities supply most members of the ASO Address Council, and the Council selects the nominees for ICANN Board Seats 9 and 10. Yet selection is indirect: ten Council members are chosen through regional community procedures, five are appointed by RIR boards, the Council votes confidentially, ICANN manages due diligence, and the Empowered Community formally designates the directors.
  • Once appointed, a director's legal duty runs to ICANN rather than to the ASO, an RIR, an employer, or a regional constituency. The ASO AC's own procedures classify ICANN directors as non-representative appointees, deny the Council ordinary removal power over them, and make updates invitational rather than mandatory.
  • A removal power exists through ICANN's Empowered Community framework, but it is an emergency constitutional mechanism, not a routine recall ballot. Petitions pass through ASO and NRO bodies, prescribed dialogue and public stages, a high approval threshold, and tightly timed notices. Regional voters therefore exercise much more practical influence at selection and reappointment than during the middle of a term.
  • The appropriate repair is not to convert directors into instructed regional delegates. It is to shorten the accountability distance through public appointment criteria, regular reasoned engagement, a clear petition intake rule, transparent reappointment review, and published explanations of how community concerns were heard without compromising board confidentiality or fiduciary judgment.

Two seats create an expectation larger than the legal relationship

The phrase "ASO seats" invites a natural inference. If the Address Supporting Organization selects the people who occupy Seats 9 and 10 on the ICANN Board, those directors appear to belong to the numbers community in the same way that an elected legislator belongs to a district. Operators may expect them to carry regional positions into the boardroom, report back after votes, defend the RIR system against institutional encroachment, and face recall if they cease to do so. The vocabulary of representation supplies the expectation before any rule is read.

The rules establish a different relationship. The ICANN Bylaws assign two Board positions to people nominated by the ASO and formally designated by the Empowered Community. They also state, in Section 7.7, that every director serves as an individual and must act in what the director reasonably believes are ICANN's best interests. A director is not the representative of the Supporting Organization that nominated the person, an employer, or another constituency. The same Bylaws require familiarity across the Internet's institutional and technical domains, not loyalty to one source of appointment.

The ASO AC Operating Procedures reinforce that separation. They define ASO-appointed ICANN Board members as "Non-Representative Appointees." They allow the Council to remove a representative appointee, but expressly say it has no power to remove a non-representative appointee. Representative appointees are expected to provide regular updates at least twice a year; non-representative appointees may merely be invited to update the Council or attend as observers.

That legal independence is not an accidental gap. A director controls the affairs of the corporation with the whole Board and assumes duties that cannot be reduced to a regional instruction. ICANN decisions can connect names, numbers, protocol parameters, budgets, corporate risk, contracts, human rights commitments, security, and the global public interest. A person who arrives through the numbers community must still evaluate the complete record. If the ASO could direct votes and dismiss the director immediately for disobedience, the Board would become a conference of envoys rather than a fiduciary body.

The accountability question therefore begins with a distinction. Independence from instructions is defensible. Independence from explanation is not automatically defensible. A sound appointment system protects directors from mandate while maintaining enough visibility, engagement, and exceptional correction to show regional communities that the nomination power is consequential after the announcement photograph.

The chain begins in five different regional systems

The ASO Address Council is also the Number Resource Organization Number Council. According to the ASO's description of the Council, it has fifteen members, three associated with each of the five RIR regions. In each region, the community selects two members through its own open and transparent procedures, while the RIR's executive board appoints one. Ten seats therefore have a direct regional-community origin and five have an institutional-board origin.

This is already more layered than a global electorate choosing a director. A voter in one region does not vote for Seat 9 or Seat 10. The voter may help select two of the three Council members associated with that region, subject to that region's eligibility rules, electorate definition, nomination method, term, turnout, and voting system. The RIR board chooses the third. Those three join twelve colleagues selected under four other regional arrangements. All fifteen then participate in a separate global appointment decision.

The arrangement has virtues. Each region has equal numerical presence on the Council. Local communities can use procedures familiar to them. A board appointee can add continuity or institutional knowledge while the two community-selected members supply a stronger participatory connection. No single large membership base can dominate the Council merely through population. The design treats the five regional systems as peers.

It also diffuses responsibility. If a regional operator dislikes a Board selection, it is not always clear which decision should be contested. Was the local Council election unrepresentative? Did the RIR board choose an unsuitable third member? Did the region's three members vote differently from one another? Did a cross-regional coalition determine the final result? The confidential appointment ballot prevents an outsider from tracing those possibilities to individual votes. Equal regional composition does not create an equal or visible causal path from each region to the selected candidate.

Terms for Council members also vary by regional procedure. The body that selects a director is therefore a rolling institution rather than a single electorate assembled for that choice. Some voters who selected a Council member may no longer be active when the Board ballot occurs. Some Council members may have joined through board appointment rather than public election. Some may leave soon after casting the ballot. The legitimacy of the Board choice rests on the standing authority of the Council, not on a contemporaneous global mandate.

This is the first accountability distance: regional entities choose or influence intermediaries, and those intermediaries later exercise independent judgment in a worldwide selection. It is representative governance, but not direct representation.

Nomination is open; decisive evaluation is concentrated

The ASO AC ICANN Board Selection Procedures divide a normal selection into nomination, comment, interview, and selection phases. Anyone may nominate any person. The nomination period must last at least sixty days. Nominees provide a certification and questionnaire, consent to due diligence, and must satisfy the eligibility and conflict requirements in the ICANN Bylaws.

Openness at the entrance is important. It prevents the formal candidate pool from being restricted to RIR executives or sitting Council members. Published timelines and candidate information give communities an opportunity to identify qualified people and submit comments. The 2024 Seat 9 election page records a multi-month sequence: nominations, public comments, interviews, deliberation, voting, due diligence, and announcement. It also shows the geographic restriction in practice: because the other ASO-selected director resided in the RIPE region, candidates from that region were ineligible for Seat 9.

The decisive evaluation, however, narrows quickly. A five-member Qualification Review Committee creates the qualified slate, with no more than two members from one region. A five-member Interview Committee conducts written and video interviews and prepares a comparative report. The full Council may observe committee work, but the public does not participate in the interviews. Candidate information and moderated comments are public; the comparative judgment that most directly shapes the ballot is an institutional product.

The Council then votes using ranked-choice instant runoff. A candidate wins after accumulating first-choice support from an absolute majority of voting members through the elimination and transfer sequence. Ballots for appointments are anonymous. Vote counts remain internal to the ASO, and only the final success is reported outside. The 2024 page identifies the selected candidate and completion of due diligence, not the distribution of preferences.

Confidentiality protects Council members from lobbying and retaliation. It encourages candid comparison of people whose professional reputations are exposed. It may also make regional bloc bargaining less attractive. But secrecy removes the most familiar accountability instrument: a voter cannot know whether a Council member supported the candidate, opposed the candidate, abstained, or ranked alternatives in a particular order. Nor can the voter test a Council election promise against the appointment ballot.

The appointment is consequently public in inputs and private at the decisive point. That can be justified, but the justification should be acknowledged. The community contributes nominations and comments; the Council owns the final judgment.

Due diligence adds a second institutional principal

After the Council selects a candidate, ICANN independently manages due diligence. The ASO procedure says the review generally takes several weeks and is intended to identify concerns in the candidate's past or submitted information that could affect service as a director. If the selected candidate fails, the candidate with the next-highest support proceeds to review.

Due diligence is a reasonable safeguard. Directors oversee a significant global corporation, handle confidential information, approve budgets and contracts, and face conflict, integrity, and competence expectations. A community appointment system should not disable basic verification. The Bylaws also impose qualifications that the ASO cannot waive, including restrictions on certain government officials and simultaneous service on Supporting Organization councils.

Yet due diligence changes the constitutional description of the choice. The ASO does not possess an unconditional power to place any selected person on the Board. It supplies a nominee through its procedure; ICANN applies a review; the Empowered Community performs formal designation. If review identifies a significant concern, the first choice may not take office. Because the sensitive basis for such a result may not be fully publishable, the regional community may see only that another candidate advanced.

This layer divides accountability between selection quality and eligibility control. The Council can be questioned for its comparative judgment, but ICANN controls the independently managed screening. ICANN can be questioned for the fairness and relevance of due diligence, but confidentiality limits the public record. The Empowered Community's designation gives the appointment legal effect, while the substantive nomination remains the ASO's.

The design is not unusual for a nonprofit board. It does mean that the phrase "elected by the ASO" compresses distinct acts. Nomination, vetting, designation, and commencement of term belong to different hands. A clear accountability account should name each hand because remedies differ. A defect in a regional Council election is not cured through ICANN due diligence. A discriminatory due-diligence outcome is not cured by another Council ballot. A late designation notice is not evidence that the regional electorate failed.

Long chains become unaccountable when every institution can point to another. They remain governable when each stage publishes its authority, decision type, reasons to the extent lawful, and route for challenge.

Seats 9 and 10 are staggered, not jointly commissioned

The Bylaws provide separate cycles for the two ASO-originated seats. Seat 9 and Seat 10 begin regular terms in different three-year sequences. The current appointment page accordingly lists two directors with different commencement and end dates. Staggering protects continuity: the numbers community does not replace both connections to the Board at once, and one experienced director may remain while another begins.

Staggering further weakens the idea of a single electoral mandate. The Council that chose one director may have a materially different membership from the Council that chose the other. Regional elections may have changed, board appointees may have rotated, and public concerns may have shifted between the two ballots. The directors are not a ticket, a caucus, or a delegation with a shared platform. They are separately nominated individuals serving concurrent portions of separate terms.

The Bylaws also require that two directors nominated by the same Supporting Organization not come from the same country or geographic region at the same time. This widens geographic distribution but constrains candidate choice. A highly regarded candidate can be ineligible because of the residence or citizenship classification used for the other seat. The restriction treats diversity across the pair as a Board value that can outweigh an unconstrained regional preference in one election.

Three-year terms, renewable within the Bylaws' consecutive-term limit, make reappointment the strongest ordinary accountability point. A serving director who seeks another term returns to comparison against other nominees, community comment, Council interviews, a new confidential ballot, and due diligence. That is real leverage. It is also delayed. A director selected shortly after one regional Council cycle can serve through substantial turnover in the communities and Council before facing a new decision.

An extraordinary selection fills a premature vacancy for the remainder of the term. The Council may draw from sufficiently supported candidates in the previous selection or use accelerated public or Council nominations. This maintains institutional continuity, but an abbreviated replacement can place even more weight on the Council's internal judgment. Regional voters do not receive a special Board ballot when a vacancy opens.

The calendar therefore defines practical influence. Communities can shape the Council continuously, contribute to candidate review during published selection periods, and affect reappointment at term boundaries. Between those moments, influence depends on engagement rather than electoral control.

The director's duty deliberately breaks the mandate chain

It would be easy to diagnose the accountability gap as a drafting failure and propose that Seats 9 and 10 must vote according to ASO instructions. That remedy would conflict with the office being filled. ICANN directors are responsible for the corporation as a whole. They may receive information from a nominating community, but they cannot outsource judgment to it.

The distinction matters most when interests diverge. An RIR community may prefer a budget outcome that protects number-resource engagement while the Board confronts a corporation-wide shortfall. The NRO may favor a contractual interpretation that preserves its institutional autonomy while ICANN counsel identifies broader legal risk. Operators may oppose a measure because of implementation cost while other parts of the Internet community identify security or rights effects. A director must hear the numbers perspective without presuming it is conclusive.

Board confidentiality creates another boundary. Directors can explain general reasoning and engagement, but they cannot disclose privileged legal advice, personnel information, security details, or confidential negotiations merely because the nominating body asks. A reporting duty that demands a transcript of deliberations would damage the Board and place the director in recurring conflict.

Independence also protects the regional communities. If a director were formally their agent, every Board vote could be attributed to five RIR communities regardless of whether those communities had considered it. Governments or litigants might treat the director's action as an authorized regional position. Minority views inside each region would disappear behind a claim of delegated consensus. The non-representative rule prevents that false unity.

But fiduciary independence should not become social isolation. A director can maintain open channels, attend public meetings, explain the factors considered, identify matters on which more numbers-community evidence is needed, and correct misunderstandings about Board authority. None of those practices requires a promised vote. The test is whether engagement informs judgment, not whether it commands the outcome.

Accountability in this setting means answerability plus lawful correction, not obedience. The present architecture secures the second only at a high constitutional threshold and leaves the first too dependent on invitation and personal practice.

The 2004 settlement and the 2016 reforms solved different problems

The institutional layers make more sense when their dates are separated. The 2004 ASO agreement replaced the earlier arrangement with one built around the Number Resource Organization and its Number Council. The settlement connected the five RIRs to ICANN while preserving regional policy development and operational autonomy. The Number Council would serve as the ASO Address Council, oversee the global policy route, advise on number allocation, and select the people for the ASO's Board positions. Its logic was coordination among established regional institutions, not direct election to a global legislature.

The appointment chain reflects that bargain. Each region contributes an equal number of Council members even though RIR memberships and service populations differ. Regional procedures retain control over the choice of community members, and RIR boards retain one appointment each. The Council then acts collectively at the global layer. A entity's authority is transformed at every step: local participation becomes Council membership; Council membership becomes a confidential global vote; the winning vote becomes a nomination to a corporate board.

The 2016 ICANN accountability reforms addressed a different concern. The transition away from United States government stewardship required enforceable community powers over ICANN's corporate center. The Empowered Community became the legal mechanism through which Supporting Organizations and Advisory Committees could reject specified decisions, inspect records, initiate community review, remove directors, or recall the Board. That reform added a removal route that the ordinary ASO appointment procedure does not contain.

It did not reverse the director's fiduciary status. On the contrary, the Bylaws continued to say that directors are individuals acting in ICANN's interests, while making them legally removable through prescribed community action. The result is a dual structure: nomination authority remains distributed through the 2004 regional coordination settlement, and extraordinary correction operates through the 2016 corporate accountability settlement.

That history explains why the route back is not the route forward in reverse. The Address Council selects because the ASO agreement assigns it expertise and appointment responsibility. The NRO EC directs the ASO's Empowered Community act because the ASO's decisional-entity procedures place that corporate power there. The two bodies are products of connected but distinct institutional purposes.

The design problem is therefore not simply that an old rule omitted recall. It is that the later constitutional remedy was added above a regional appointment system without creating a comparably clear layer of routine answerability between selection and crisis. The formal gap narrowed; the everyday accountability distance remained.

Routine reporting is weaker than the appointment rhetoric suggests

The sharpest textual evidence of the gap lies in Section 8 of the ASO AC Operating Procedures. Representative appointees are expected to update the Council at least twice annually. Non-representative appointees may be invited to update and may be invited to observe. Because ASO-appointed ICANN directors are explicitly non-representative, the mandatory expectation does not attach to them.

That rule is legally coherent. A Board director should not be managed as an emissary. Yet an update is not necessarily an instruction. A regular public appearance can explain the Board's agenda, describe how the director seeks input, and surface questions without promising to reveal confidential deliberation. Treating every reporting expectation as a threat to independence sacrifices a low-cost accountability tool.

The practical result can vary by officeholder. One director may attend RIR meetings, participate in Council calls when invited, publish accessible reflections, and seek evidence before important Board discussions. Another may comply fully with ICANN duties while maintaining little visible connection to the communities that supplied the appointment path. Both can satisfy the formal non-representative model. The difference is reputational practice rather than enforceable baseline.

This variability makes reappointment harder to evaluate. Public candidate pages document statements before selection. Board minutes and resolutions show institutional outcomes. They do not necessarily reveal whether a director maintained a serious two-way channel with the numbers community, raised relevant technical consequences, distinguished personal judgment from community input, or responded when regional concerns conflicted.

A regular accountability statement could improve the record without creating delegation. It might list meetings attended, topics on which input was sought, published Board materials relevant to number resources, declared conflicts, and unresolved questions. It should explicitly say that the director did not receive or accept binding instructions. Sensitive Board information would remain protected. The value would be continuity of engagement and a comparable record at reappointment.

The Council could likewise publish an annual collective assessment of whether the engagement arrangements worked, without scoring a director's votes. It could identify unanswered requests, missing channels, or subjects requiring earlier consultation. Accountability would then focus on the quality of the interface rather than policy conformity.

Removal exists, but not as a direct regional recall

Since the 2016 accountability reforms, ICANN's Empowered Community has possessed the power to remove individual directors other than the President and to recall the Board. Annex D of the Bylaws permits an individual to submit a petition to the applicable Decisional Entity seeking removal of a director nominated by that Supporting Organization or the At-Large Community. For an ASO-nominated director, the applicable entity is the ASO.

The procedure is intentionally serious. The applicable entity must decide whether to accept the petition within a defined period and invite the affected director and Board leadership to dialogue with the petitioner and the entity's representative. If the petition proceeds, a community forum and a public comment stage follow. The final removal notice requires support within the applicable entity at the three-quarters level determined under its own procedures. A failed process can protect the director from another removal attempt for the remainder of the term.

The ASO's procedures for Empowered Community powers route the matter through the NRO Secretariat and NRO Executive Council. If an ASO-originated director is targeted, the ASO AC is informed and may make a recommendation, but the NRO EC decides how the ASO acts. The ASO's Empowered Community explanation says the NRO Chair acts as the ASO decisional representative only by consensus of all five NRO EC members. In other words, the body that selected the director advises; the executive heads of the five RIRs control the formal ASO act.

This route is a meaningful legal power. It prevents a director from being literally untouchable until term end. It also differs radically from a constituency recall. A regional voter cannot file a local ballot and remove the director by majority. A petition enters a global institutional channel, faces threshold and timing requirements, engages the director and Board leadership, and must achieve a level of support designed for exceptional failure.

The difference is appropriate up to a point. Removal should not punish a director for one unpopular vote or for declining an instruction the Bylaws prohibit. It should be available for grave misconduct, persistent incapacity, undisclosed conflicts, serious breach of duty, or a collapse of confidence supported by evidence. The danger is not the high final threshold alone. It is uncertainty about how ordinary entities get a credible allegation to the threshold.

The ASO AC and NRO EC occupy different sides of the loop

Appointment and removal do not mirror each other. The ASO AC defines and runs the selection procedure, evaluates candidates, and chooses the nominee. In an Empowered Community removal action, the ASO AC may consider the petition and recommend an approach, but the NRO EC is the decisional body directing the ASO representative.

This asymmetry can be defended as a check. The Council that chose a director may be reluctant to admit error or may have become hostile over a legitimate disagreement. Requiring a separate executive body to act prevents an appointment majority from converting directly into a dismissal majority. Consensus among all five RIR chief executives ensures that removal cannot be driven by one region.

It can also weaken democratic traceability. Five RIR executives include officers accountable through five different corporate governance systems, not one global electorate. Their consensus may be prudent, but regional entities may have no common mechanism to direct or review it. The ASO AC's recommendation can reflect a fifteen-member body with ten community-selected members, yet it is not controlling. A unanimous NRO EC position can override or proceed without timely Council input under the published description of the role.

The structure is particularly sensitive when the alleged problem concerns RIR institutional interests. If a director challenges an NRO position in what the director believes to be ICANN's best interests, RIR executives should not be able to characterize independence itself as disloyalty. Conversely, if the director has an undisclosed conflict involving an RIR, executive knowledge may be essential. Neither body is inherently the neutral one.

The answer is a reasoned division of tasks. The AC should evaluate appointment expectations, public engagement, and the evidence bearing on individual fitness. The NRO EC should evaluate the ASO's legal and institutional act, declare conflicts, and explain how each region was consulted. The final rationale should distinguish misconduct from policy disagreement and identify the evidence standard applied. If the bodies differ, both positions should be published unless lawful confidentiality prevents it.

Without that record, the same separation that protects a director can make regional dissatisfaction disappear between institutions. A feedback loop is not complete merely because every node received an email.

Regional voters have leverage, but it is front-loaded

It would be inaccurate to say regional voters have no influence. They choose ten of the fifteen people who compose the Address Council. They can nominate Board candidates, comment during the published phase, question Council candidates about selection philosophy, replace Council members at later regional elections, and argue against a director's reappointment. They can also submit a removal petition under the applicable rules.

The issue is timing and conversion. Influence is strongest before a director takes office and when a term returns to selection. During the term, a voter cannot see the appointment ballot, cannot instruct the director, cannot require a report under the representative-appointee rule, and cannot use a simple regional recall. A concern must be converted into Council attention, NRO EC action, or reputational pressure. Each conversion adds time and an institutional judgment about whether the concern deserves escalation.

That front-loading affects whose voices matter. Well-connected entities who attend several RIR and ICANN meetings can maintain informal contact with directors and Council members. Smaller operators, civil society researchers, and entities working in less-resourced language communities depend more heavily on formal notices and published records. If the formal channel is thin, access becomes a substitute for accountability.

Regional turnover can further break the loop. A entity who voted for an AC member may expect that person to scrutinize a Board appointee, but the AC member's term can end before the issue matures. The successor may not know the selection rationale because comparative reports and vote counts are confidential. Institutional memory resides with the Secretariat and long-serving entities rather than with a public decision file.

The practical accountability instrument is therefore reappointment. It should be treated as an evaluation of completed service, not a presumption in favor of continuity. Incumbents bring valuable Board experience, but that experience can dominate comparison if challengers lack access to equivalent information. A reappointment file should address attendance, conflicts, committee work, engagement, reasoned public communication, and the contribution of number-resource expertise, while respecting confidential Board material.

Front-loaded influence is not illegitimate. It becomes weak when the system pretends that an appointment automatically produces continuous representation.

Confidential ballots conceal coalition, not just individual choice

The appointment rule keeps individual votes and counts inside the ASO. This prevents public attribution of preferences, but it also conceals aggregate information that could be released without identifying a ballot. Communities do not learn whether the final choice was overwhelming or narrowly achieved after several transfers. They cannot see turnout beyond the procedural rule, the number of abstentions, or whether the selected candidate began as the leading first choice.

Aggregate information matters because selection legitimacy is not binary. A candidate chosen with broad cross-regional confidence enters office differently from one who survived a fragmented contest by the smallest permissible margin. Both are lawfully selected, and neither should be treated as a regional delegate. But the Council might engage more deliberately after a close result, explain the comparative criteria, or improve the next candidate search.

Publishing individual ballots would likely do more harm than good. Council members could face employer or regional pressure, candidates could target perceived opponents, and future votes could harden into blocs. Publishing de-identified round totals, turnout, and abstentions would preserve ballot secrecy while revealing the strength and shape of collective support. Even that approach requires care in a fifteen-member body, where regional statements might make identities inferable.

At minimum, the Council can publish a reasoned selection statement. It should identify the capabilities sought, the evidence considered, the treatment of public comments, the diversity constraint, and the reasons the selected candidate best met the role. It need not compare named unsuccessful candidates or reveal confidential interviews. Present practice announces who won; accountability requires some explanation of why the institution exercised its judgment that way.

The same principle applies to reappointment. "Selected" should not be the only public finding after three years of observable service. A statement can explain why continuity, Board contribution, engagement, independence, and expertise justified another term. If the incumbent loses, the institution need not publish adverse personal material, but it can explain the qualities prioritized in the successor.

Secrecy should protect deliberation. It should not erase institutional reasons.

Geographic diversity is real, but it is not regional instruction

The two ASO-originated directors cannot come from the same ICANN geographic region at the same time. The rule creates visible geographic distribution, and the Council's own composition gives every RIR region three entities. These are meaningful structural safeguards against concentration.

Geography does not prove that a director carries a region's policy. A person residing in Africa may have worked globally, belong to no RIR governing body, and disagree with many African operators. A director from the RIPE region may understand address policy deeply without expressing the preference of every network from Europe, Central Asia, and the Middle East. The Bylaws' geographic categories are coarse, and the RIR service regions do not map perfectly onto political identity.

Confusing diversity with delegation creates two errors. First, it loads one person with the impossible task of representing a vast and internally divided region. Second, it lets the institution claim regional accountability merely because a seat is occupied by someone who satisfies a location rule. Descriptive presence can improve deliberation, but it does not replace consultation or explanation.

Selection should therefore examine more than residence. The Council can ask how candidates seek views beyond their professional circle, handle linguistic and resource barriers, recognize differences among RIR communities, and distinguish regional experience from authority to speak for a region. Public comments can reveal whether a candidate has trust across institutional boundaries, though popularity should not displace competence and independence.

After appointment, engagement plans can be geographically balanced without assigning mandates. The two directors might coordinate coverage of RIR meetings while making clear that neither speaks for a region. Remote sessions and written questions can reach entities who cannot travel. Published summaries can record issues heard rather than positions promised.

The accountability loop improves when geography determines access opportunities, not presumed obedience.

Three failure scenarios show what ordinary rules must do

Consider first a director who supports a Board decision opposed by every RIR executive. The director explains that confidential legal and financial evidence changed the balance and publishes the non-confidential factors. There is no conflict, misconduct, or neglect. Removal would be inappropriate. The appointment system selected an independent fiduciary, not a fifth signature for the NRO. The proper response is debate and later reappointment review.

Now consider a director who repeatedly misses Board work, fails to disclose a material relationship, and stops responding to reasonable requests for engagement. Those facts concern fitness and duty rather than one policy position. The Council should be able to document the pattern, ask for a response, and, if the evidence warrants, recommend that the NRO EC accept a removal petition. A clear intake rule would keep the allegation from depending on personal access to an executive.

In a third case, regional communities broadly lose confidence because the director appears to dismiss number-resource consequences, but no single act constitutes misconduct. A constitutional removal process may be too severe, especially if the director otherwise fulfills fiduciary duties. Waiting silently for term end is also poor governance. A structured public engagement, independent performance review against the published appointment criteria, and an explicit statement about reappointment can address the confidence problem without pretending there is cause for immediate dismissal.

These cases require different tools. Policy disagreement needs explanation. Fitness failure may justify escalation. Relationship failure needs engagement and prospective electoral consequence. A system with only confidential selection and extraordinary removal forces all dissatisfaction into either gossip or a constitutional petition.

The Council should publish a classification framework before a controversy. It can state examples of issues normally resolved by engagement, issues relevant to reappointment, and issues capable of supporting removal. The framework must preserve judgment and avoid creating contractual tenure rights, but it would help petitioners supply relevant evidence and deter campaigns based solely on an unpopular vote.

Predictability protects both communities and directors. A director should know that independent judgment is safe and that concealment, incapacity, or serious breach is not.

A stronger loop can preserve fiduciary independence

Reform should begin at selection. Before nominations open, the Council should publish a role profile distinguishing Board-wide fiduciary duty from the value of number-resource expertise. It should list the skills sought in that cycle, the expected public engagement baseline, conflict expectations, time commitment, and the evidence that will inform reappointment. Candidates would then consent to answerability without promising instructed votes.

During selection, the Council should publish a reasoned institutional report after due diligence. The report can describe the process, public participation, evaluation dimensions, geographic constraint, and basis for the final choice. De-identified aggregate ballot information should be released where it can be done without exposing individual preferences. Candidate privacy should protect personal material, not the Council's criteria.

During the term, each ASO-originated director should offer at least two public accountability engagements a year, scheduled across RIR regions and accessible remotely. The sessions should cover Board matters relevant to Internet numbers, upcoming decisions, requests for evidence, conflicts, and the boundary between public explanation and Board confidentiality. They should not require the director to seek permission for a vote.

The Council should maintain a public correspondence register for substantive questions, showing date, topic, disposition, and whether a response was public, confidential, pending, or outside the director's role. This would expose neglected issues without publishing sensitive content. Regional communities could see that their concerns entered the interface even when the outcome differed.

For removal, the ASO should publish a plain-language petition form, evidence standard, filing address, conflict procedure, and exact relationship among the petitioner, ASO AC, NRO EC, and Empowered Community notices. The current legal texts describe the stages, but ordinary entities should not need to reconstruct the route across ICANN Annex D and several ASO pages. Timely publication should include reasons for acceptance, rejection, support, objection, or abstention.

At reappointment, the incumbent should face an express service review against the same public criteria as challengers. Board experience is an asset, not an entitlement. The Council's final statement should show what it learned from the preceding term and why another appointment serves the Board.

None of these measures turns a director into a delegate. They make independence observable and accountable.

The test is whether concern can travel back through the chain

The ASO appointment architecture has a clear forward path. Regional communities and RIR boards populate the Council. The Council opens nominations, receives comments, interviews candidates, votes, and sends a name for due diligence and formal designation. The director begins a staggered term and serves as an individual fiduciary. Every step has an identified institution.

The reverse path is less clear. A regional entity with a concern must decide whether to contact the director, a local Council member, the ASO Secretariat, an RIR executive, or the NRO Secretariat. The concern may be treated as policy disagreement, a request for information, a reappointment factor, or a removal petition. The ASO AC can advise on removal but cannot use its ordinary appointment power to end the term. The NRO EC directs the ASO's Empowered Community act. The formal threshold is high, and routine reporting is optional.

That asymmetry is the accountability loop. Forward authority is specific; backward influence is dispersed. Some of the asymmetry is necessary because fiduciary directors cannot be permanently tethered to electors. Too much of it leaves the appointment power symbolically impressive and operationally remote.

The relevant questions are practical. Can a regional voter identify which Council members exercise the appointment judgment? Can the public understand why a candidate was selected without seeing private ballots? Can entities obtain regular explanations without asking for confidential Board deliberations? Can a serious allegation enter a known petition channel? Can the Council and NRO EC show how they divided responsibility? Can reappointment distinguish service from incumbency?

If those answers are yes, Seats 9 and 10 can combine independent Board judgment with a credible numbers-community origin. If they are no, the formal right to nominate two directors will continue to exceed the community's ability to evaluate the authority it helped create.

The objective is not recall on demand. It is a chain in which appointment, engagement, correction, and renewal are all visible enough to deserve confidence.