Summary

  • ICANN's Ombuds can obtain necessary information, assess evidence, mediate, investigate unfairness or harassment, make findings, recommend corrective action, report serious unfairness and contribute a public evaluation to qualifying Reconsideration Requests. These powers can resolve many failures before positions harden.
  • The office cannot by itself make, change or set aside a policy or Board decision, reverse an act, award compensation, bind the ordinary respondent or supply the interim relief available in independent review. Even a harassment remedy refused by a respondent goes to the Board for any imposed action.
  • The 2026 Framework adds clearer case categories and response times, while FY25 reporting shows that most matters were collaborative or self-managed. The next accountability test is recommendation uptake: every formal recommendation needs a public response, implementation owner, completion evidence and a route to enforceable review when voluntary correction fails.

A complaint office is not a small court

The word ombudsman carries several traditions. In some legal systems it describes a public officer with statutory investigatory powers and a route to Parliament. In organizations it often describes a confidential, impartial resource that helps people navigate conflict and administrative unfairness. ICANN's office belongs mainly to the second tradition, with several duties written into the corporation's Bylaws.

ICANN now generally calls it the Office of Ombuds and its office-holder the Ombuds, while Article 5 of the current Bylaws retains “Ombudsman.” The change in usage does not alter the central allocation of authority. The Ombuds evaluates and seeks resolution. The Board, a formal accountability panel, a court or another authorized body makes any binding decision available under the applicable route.

This is not a defect accidentally omitted from the charter. The 2026 Ombuds Office Framework and Process states directly that the office has no unilateral power to make, change or set aside a policy, administrative decision or Board decision, and cannot reverse an act or failure to act. It can investigate the events and use collaborative methods to seek resolution.

The difference between investigation and order is the foundation of the office. An investigator can establish facts, identify a broken rule and recommend repair. An order-maker can compel a consequence. A mediator can help parties find an agreement. An adjudicator can decide despite disagreement. An advocate for fairness can expose an institutional blind spot. A rights-holder can demand compliance.

Collapsing these roles creates two dangers. A complainant may rely on the Ombuds when a formal filing deadline or urgent stay is needed, only to learn that contact did not preserve the claim. ICANN may cite the existence of the office as proof that a remedy exists even when the requested relief was beyond its power. Accurate role definition protects both the office and the people who approach it.

The right question is therefore not whether an ombuds office without orders is weak. It is what classes of problem can be corrected through information, persuasion, facilitated agreement and exposure; what evidence shows those corrections occurred; and where an enforceable route must begin.

The charter is built around fairness and early resolution

Article 5 makes the principal function an independent evaluation of complaints from community members who believe that ICANN staff, the Board or a constituent body treated them unfairly. The Ombuds is to serve as an objective advocate for fairness, clarify issues and, where possible, resolve complaints through negotiation, facilitation and shuttle diplomacy.

The phrase “advocate for fairness” does not mean advocate for the complainant. Neutrality and independence are express duties. The Ombuds should not assume that the person bringing a complaint is right, nor defend the challenged institution. The office advances fair treatment by examining evidence and making the route to resolution intelligible.

The jurisdiction is broad in actor and narrow in subject. Community members may raise specific actions or failures to act by staff or the Board, and the office may address conduct by constituent bodies. The current framework also covers complaints under the Community Anti-Harassment Policy and Expected Standards of Behavior.

Several subjects remain outside authority: ICANN administrative matters concerning its own organization, personnel and employment issues, Board membership questions, and vendor or supplier relations. The Ombuds may decline complaints that arrive too late, lack sufficient personal interest, are repetitive or abusive, lack credibility, have an adequate alternative route, or cannot usefully be resolved through the office.

The office also ordinarily waits when a dispute is premature. If the relevant group has not completed its decision, has not had time to answer or can still repair the matter itself, the Ombuds may refer the complainant back. That discipline prevents every disagreement from becoming a quasi-appeal before the first decision-maker has acted.

Early resolution is not merely a cheaper version of judgment. It addresses a different stage of conflict. A missed response, misunderstood rule, hostile exchange or incomplete record may be corrected before it becomes an entrenched legal dispute. The Ombuds can ask the right person to answer, help the parties define their interests and suggest a proportionate path.

The limit is equally clear. Once the question is whether a valid Board decision must be stayed or invalidated despite ICANN's opposition, early-resolution authority has run out. The complainant needs a mechanism with power over the disputed act.

Information is the office's strongest non-coercive asset

An ombuds office cannot investigate credibly if the organization controls every relevant record. Article 5 therefore gives the Ombuds access to necessary information and records from staff and constituent bodies, subject to confidentiality. The 2026 Framework describes access to relevant and necessary information from the Board, staff and constituent bodies for evaluation and resolution.

This access can correct several asymmetries. A community volunteer may know that a selection felt irregular but not know who participated. A complainant may have one email while the decision-maker holds the full exchange. A chair may believe a deadline was met when another office has contradictory records. By assembling a shared factual account, the Ombuds can turn accusation into a specific question.

The right is not a publication right. Confidential material remains protected, and parties can designate nonpublic communications as confidential. The office can reach an informed view without making every underlying document available to the complainant or public. That distinction protects candid participation but limits external verification.

Information access also lacks an ordinary subpoena mechanism. The current Framework sets response expectations: seven days for information or an interview, extendable to fourteen and further in exceptional circumstances. If a response is missing or unjustifiably delayed, the Ombuds may proceed on the available record and can notify the Board or the President and CEO of noncompliance.

That escalation matters, but it remains different from contempt power. The Ombuds can expose a refusal and recommend a consequence. The office does not itself fine the nonresponsive party. Where role or participation restrictions are proposed, fair notice is required and the institution with authority must impose them.

The distinction should be reflected in reports. “Information obtained” is not the same as “all requested information produced.” A fair annual account would identify requests, timely responses, extensions, incomplete responses, refusals, escalation and effect. Without those measures, the public cannot know whether access works in difficult cases or only when cooperation is already likely.

Mediation corrects relationships that orders often cannot

Many ICANN disputes are not requests to invalidate a Board resolution. They concern volunteer working relationships, leadership communication, disrespect, exclusion, misunderstandings and persistent low-level conflict. A formal ruling can decide who breached a rule while leaving the group unable to work together.

The Ombuds can use mediation, shuttle negotiation, conciliation and dialogue. In direct mediation, parties may hear each other's account and design an agreement. In shuttle work, the Ombuds carries proposals between people who cannot yet meet productively. Coaching can help a community member clarify an objective, prepare a difficult conversation or respond without escalating the conflict.

These methods can produce remedies more tailored than an order. A chair may agree to publish criteria, a entity may apologize, a group may set communication rules, a selection body may give feedback, or parties may agree on how future meetings will be moderated. Because the parties help design the result, compliance can be more durable than obedience imposed from outside.

Voluntariness is both the source of that value and the ceiling. Mediation depends on willingness to participate and compromise. A powerful respondent may calculate that no agreement is preferable. A complainant may reasonably refuse dialogue where the harm is severe, the power imbalance is large or an authoritative finding is needed. Confidential settlement can also conceal a recurring institutional problem.

The Ombuds should therefore screen for suitability rather than treating mediation as the default mark of success. Relevant factors include safety, retaliation risk, urgency, repeat behavior, legal rights, need for precedent, information inequality and whether a party can negotiate freely. A voluntary method is not fair merely because both names appear on a meeting invitation.

Where mediation succeeds, reporting should distinguish the outcome. Was there an agreement, a restored relationship, a corrected decision, an answer, an apology or only a conversation? Did the remedy last? Confidentiality may prevent detail, but aggregated categories can show whether collaborative work changes conduct.

Where it fails, the office must point toward the appropriate review route. The 2026 Framework requires the Ombuds to advise complainants of applicable procedures when collaborative resolution is exhausted or unlikely to succeed. That referral is not administrative tidiness; it is the bridge between soft accountability and rights.

Procedural fairness is often correctable without deciding the merits

The 2026 Framework separates procedural fairness from interactional or relational fairness. Procedural fairness asks whether the agreed rules and standards were followed consistently. Relational fairness asks whether people were treated with dignity, respect and expected behavior during the decision.

This distinction gives the office useful room. The Ombuds need not decide who deserved a leadership role in order to find that criteria were unpublished, one candidate was denied a chance to respond or a chair failed to answer. It need not rewrite a policy to identify that the policy was applied differently to similarly placed entities.

The Framework says that in most cases the office will focus on the decision route and treatment rather than reassessing the decision itself. For a sanction such as exclusion from a community group, the Ombuds examines notice, opportunity to answer and whether objective facts supported the action. These are administrable questions even when the final merits remain contested.

Corrective action can then be modest and effective. The decision-maker can supply reasons, reopen a hearing, correct a factual error, apply the stated criteria, obtain an unbiased reviewer or repeat the decision under a fair procedure. None of these steps requires the Ombuds to become the substantive decision-maker.

The office also protects respondents. Draft findings are shared for comment on factual accuracy. Both sides can supply evidence, correct interview notes and address preliminary views. A fairness mechanism that disregards the respondent's opportunity to be heard would reproduce the defect it is meant to correct.

Still, a recommendation to rerun a decision is not the same as an order to rerun it. If the decision-maker refuses, the office can report, escalate and create reputational cost. Unless another rule makes the remedy compulsory, the original decision remains in force.

This is where soft accountability is sometimes oversold. A reasoned finding may be persuasive enough in an institution that values legitimacy. It may be ineffective where the disputed decision confers a scarce benefit, protects an incumbent or creates irreversible consequences. The more a respondent gains from refusal, the less confidence should be placed in persuasion alone.

Harassment reveals the boundary between recommendation and sanction

The Community Anti-Harassment Policy gives the Ombuds a more defined investigative role. The office can assess evidence, interview relevant people, determine whether a complaint is substantiated, substantiated in part or not substantiated, and suggest remedial action based on severity, frequency, impact, prior complaints and willingness to repair.

This is more than informal advice. A documented finding can validate harm, protect a respondent from an unsupported accusation, reveal repeat behavior and guide a proportionate response. The office's independence and confidentiality may make people more willing to report conduct than they would be through a public disciplinary proceeding.

But the remedy architecture preserves the core limit. If a respondent refuses the proposed remedial action, the Ombuds refers the matter to the Board with findings and recommendations. The Board evaluates the record and decides whether to impose the recommended or an alternative action. The coercive act belongs to the Board.

That separation can protect due process. A neutral investigator develops the facts, while the authorized institution decides a sanction. It can also weaken independence because the Board is the office's appointing and accountability body and may itself be implicated in the wider institutional context. Clear recusal, conflict and publication rules are necessary.

Confidentiality makes public evaluation especially difficult. The community should not receive identifying details that expose complainants, witnesses or respondents. It still needs aggregated evidence: number and type of complaints, time to resolution, findings, recommended remedy categories, refusal, Board referral, Board response, implementation and recurrence.

Low complaint volume is ambiguous. It could indicate little harassment, effective prevention, lack of awareness, fear of retaliation or lack of confidence in the office. The FY25 Annual Report explicitly recognizes this uncertainty. Counts cannot be interpreted without trust and accessibility measures.

The office should also state plainly that contact is not notice to ICANN of a legal action or cause of action. A person who must make a formal report under another rule should not assume that a confidential Ombuds conversation completed that duty. Safety and rights depend on route clarity.

Recommendation power works by changing the audience for refusal

A private complaint can be easy for a powerful institution to dismiss. An Ombuds recommendation changes the audience. The disputed conduct is no longer described only by the affected person; an independent office has reviewed evidence and stated what correction is warranted. Refusal then requires the institution to oppose both the complainant and the office's reasoned assessment.

The 2026 Framework lets the Ombuds recommend action to the Board and include recommendations in annual reporting to reduce future complaints. It also creates response expectations. A final report or recommendation should receive a substantive response in sixty days, or ninety days with a suitable reason. Further extension requires a written explanation requested before the first deadline expires.

These timelines are a substantial improvement over an open-ended request. Delay can no longer be presented as invisible consideration. The Ombuds can identify nonresponse and escalate it. A written rationale can show whether the recipient accepts the finding, rejects the facts, disputes authority, proposes another remedy or needs more time.

Yet the wording remains soft. The Framework repeatedly uses “should” and “endeavor.” It does not convert every recommendation into an enforceable command after day ninety. Escalation creates attention; it does not automatically change the underlying decision.

Recommendation uptake must therefore be measured as a chain. The first question is whether the recipient answered. The second is whether the answer accepted, partially accepted, substituted or rejected the recommendation. The third is whether an owner and date were assigned. The fourth is whether the action was completed. The fifth is whether the correction had the intended effect and prevented recurrence.

Annual reports have historically emphasized complaint volume, subject matter and broad outcomes. Those are necessary but limited public evidence. “Resolved” may mean the parties agreed, the person received information, the complainant withdrew or no further action was possible. It does not necessarily mean the disputed conduct changed.

A recommendation register can protect confidentiality by using anonymized identifiers and remedy categories. It can state that a community group was advised to publish selection criteria, that the group accepted, that criteria appeared by a date and that no personal details are disclosed. For sensitive cases, even the category may need delay or aggregation, but the uptake denominator should not disappear.

The office's authority will grow when acceptance is not the only visible outcome. Publishing reasoned rejection proves that the recipient was free to disagree and makes the boundary of soft power honest. Silent disappearance is more damaging than a defensible refusal.

Confidentiality creates trust and an evidence problem

People approach an Ombuds because they need a place where uncertainty, fear and anger can be discussed without immediate public exposure. Article 5 requires ICANN entities to respect the office's confidentiality determinations. The Framework extends protection to complainants, respondents, witnesses and others contacted during a case.

Confidentiality enables candid fact-finding and settlement. It can protect a volunteer who fears losing funding or opportunity, a respondent facing an unproven allegation and a witness caught between community factions. It can create the psychological safety required for an apology or admission.

It also makes institutional performance hard to verify. Outsiders cannot normally inspect the file, evaluate the evidence or know whether similarly situated complainants received similar treatment. A confidential resolution may leave the wider community unaware that a rule failed. Repeat respondents may benefit if earlier cases cannot be connected.

The answer is not to publish case files. It is to design layered reporting. Individual parties receive enough information to understand the outcome. The Board receives what it needs for oversight or action, subject to conflict safeguards. The community receives anonymized case categories, time measures, outcomes, recommendations and systemic observations. Independent reviewers receive protected access sufficient to test consistency.

The Ombuds Office accountability page says reports are anonymized and intended to share activity, trends and ideas while demonstrating that concerns receive fair and robust handling. That statement should become a measurable publication standard. Reports should explain suppression rules, category changes and the limits of small samples.

The FY25 report, for example, says the office will not report fewer than four cases in a category where privacy protection is required. That is a sensible boundary. It also means the absence of a published category cannot be read as zero cases. Every table needs that caveat.

The Ombuds may also issue a report on a particular matter and ordinarily post it unless publication would be inappropriate. This is a valuable systemic tool because a pattern can be exposed without turning every complaint into a public contest. The decision not to publish, however, is largely entrusted to the office. Periodic independent review should examine whether that discretion is used consistently.

Transparency and confidentiality are not opposites. Good transparency explains what cannot be disclosed, why, who reviewed the restriction and what aggregate evidence remains. Poor transparency invokes confidentiality as an endpoint.

Independence is protected structurally and doubted socially

The Ombuds operates within ICANN while examining conduct by ICANN institutions. The Bylaws protect the office through a full-time post, a budget submission that the President must pass to the Board unchanged, and a requirement of three-quarters of the full Board for dismissal. The office has a right to communicate with the community without obstruction.

These protections are meaningful. A manager cannot casually reduce a proposed budget before it reaches the Board or dismiss the office-holder after an inconvenient report. Information access is grounded in the Bylaws rather than ordinary managerial permission.

Dependence remains. The Board appoints the Ombudsman, establishes the final budget and serves as the formal accountability point. The office may report serious unfairness to the same Board and needs a Board committee's authorization to open an own-initiative investigation. Structural protection reduces pressure; it cannot remove every perception of it.

The 2017 independent assessment of the office identified needs around role clarity, standing, independence, transparency and non-complaint work. The resulting Work Stream 2 recommendations proposed clearer case procedures, response deadlines, mediation capability, diverse staffing, longer fixed terms, communications measures and an Ombuds Advisory Panel.

The official implementation page shows uneven progress. Formal mediation capability, diversity measures, stronger contractual terms, strategic focus and communications work are marked complete. The advisory panel remained not started on the latest page reviewed, while several procedural items were still shown as in progress even though the Board approved a revised Framework in June 2026. That documentary lag itself deserves correction because users need to know which protections are operative.

Perception data confirms the issue. In a May 2025 baseline survey reported by the office, 80 percent of respondents who had previously contacted it said they would recommend it to someone else, but only 40 percent of respondents believed it operated independently from other parts of ICANN. These figures have unknown response and selection limits, yet the gap is instructive: service users can value the office while the wider institution doubts its distance from power.

Independence is proved through behavior. The office should publish appointment criteria, term protections, budget requests and final allocations, recusals, use of independent counsel, delayed responses, rejected recommendations and cases in which it criticized the Board. An assertion of neutrality cannot substitute for evidence of institutional resistance.

An advisory panel could help with selection, performance oversight and periodic external review if it remains outside case decisions. It should not receive confidential case guidance or become a shadow appeal body. Its value would be to reduce the Board's exclusive control over the conditions of office, not to dilute the Ombuds' own responsibility.

The FY25 figures describe a service of navigation more than adjudication

The FY25 Annual Report provides the clearest recent portrait. It records 36 in-scope cases: 16 classified as unfairness and 20 as interpersonal or relational. It also records 212 out-of-scope submissions. Approximately 15 percent of all submissions were in scope.

That ratio should not be treated as a rejection rate without context. Many people contact ICANN about domain registration, abuse, contract compliance or other matters beyond the Ombuds mandate. Referral can itself be useful. But the number also shows persistent confusion about what the office can correct.

Among in-scope cases, 78 percent involved information or collaborative problem-solving and 22 percent used a complaint route. Seventy percent of closed cases were classified as self-managed: the Ombuds supported an individual to take action without requiring the person to report back. Ninety-four percent of in-scope cases were closed during the period, while three collaborative cases remained open.

These figures match the office's design. Most value comes from advice, navigation, coaching and voluntary resolution rather than formal findings. People are using it as an early intervention service.

They also reveal an evidence gap. A self-managed case may be a success because the person gained confidence and solved the problem. It may also end without correction, and the office may never know. Closure is an administrative state, not proof of remedy.

The report identifies group dynamics, incivility and procedural unfairness as leading issues. It describes complaints about unanswered questions, opaque decision influence, cliques, unequal access to opportunities and weak feedback in leadership selection. These observations give the office systemic value: separate confidential stories can reveal a pattern that no single complainant could prove.

The suggested responses—better leadership support, wider candidate pools, stronger selection criteria, inclusive leadership and closed feedback loops—are sensible. They remain suggestions. The report does not provide a complete adoption and effect table for each community group.

Future reports should preserve the human account while adding remedy evidence. For collaborative cases, report agreement and follow-up rates where consent permits. For self-managed cases, offer voluntary outcome checks. For recommendations, track acceptance and completion. For referrals, state whether the person reached the correct destination. Data should illuminate correction rather than reward case closure.

The 2015 own-initiative inquiry shows reach and dependence at once

The 2015 Community Priority Evaluation inquiry is an instructive public example. The Ombudsman had received multiple complaints about evaluation of community applications in the new generic top-level-domain program and sought authority from the Board Governance Committee to investigate the issue on his own initiative.

The committee approved. The Ombudsman reviewed procedure, spoke with ICANN staff and the Economist Intelligence Unit, invited submissions and assessed allegations about cost, opacity, evaluation and fairness. The resulting report criticized aspects such as panel anonymity and clarity of fees, but did not find problems serious enough to recommend changing completed applications. It suggested greater openness for a later round.

The case demonstrates the office's ability to aggregate complaints and examine a systemic question. A single applicant might focus on its score; an own-initiative inquiry can look across cases, identify common concerns and distinguish error from structural unfairness.

It also demonstrates limits. The Ombudsman needed Board committee authorization to start. The inquiry could investigate and recommend, not redo evaluations. Its conclusion was contestable and not an order. Any design change depended on later decision-makers.

The report's public reasoning was therefore the remedy in part. It described the allegations, evidence and conclusion, creating a record others could challenge or use in later policy work. Even without changing an application, such a report can reduce uncertainty and identify reform questions.

But publication alone should not be counted as correction. The proper follow-up is to ask whether later applicant rules changed panel disclosure, fee clarity or evaluation safeguards and whether those changes addressed the concerns. A report that ends at recommendation produces institutional memory only if someone maintains the link to implementation.

Own-initiative authority is especially useful where fear, fragmentation or lack of standing prevents a single complaint. Requiring prior Board approval protects against an office roaming beyond its remit, but it can chill investigation of a pattern involving the Board. A narrower safeguard would permit the Ombuds to begin within clearly defined jurisdiction and require independent review of scope only when challenged.

Until that changes, each request and Board response should be public unless confidentiality makes disclosure unsafe. A denied request would be critical accountability evidence. Permission should not disappear into committee silence.

Reconsideration gives the Ombuds a formal but nonfinal role

Article 4 assigns the Ombuds a role in Reconsideration Requests that survive summary dismissal. The office reviews the request, may obtain outside expertise within budget, and provides the Board Accountability Mechanisms Committee with a substantive evaluation. Under the current Bylaws that evaluation is due within fifteen days.

This is distinct from ordinary confidential complaint work. The evaluation becomes part of a formal accountability proceeding and is made public. The Ombuds contributes independent analysis, but the committee and Board retain decision authority. If the Ombuds previously took a position on the matter or impartiality could reasonably be questioned, recusal is required.

The arrangement can improve reconsideration by placing a fairness specialist between the requester and the Board committee. It can expose overlooked information, procedural inconsistency or an inadequate response. Because the evaluation is public, the Board must engage with an independent account rather than deciding on a private staff summary alone.

It does not make reconsideration adjudication by the Ombuds. The office cannot grant the request. It cannot compel the committee to adopt its evaluation. A strong recommendation may be rejected with reasons.

FY25 illustrates the gate. Two Reconsideration Requests were filed, and both were summarily dismissed as insufficiently stated, so neither reached the Ombuds for evaluation. This is not evidence that the office failed. It shows that the Ombuds' formal role depends on an earlier committee determination.

The gate should be measured. Annual reporting should identify requests filed, summarily dismissed, referred, recused, evaluated, agreed with by the committee, accepted by the Board and implemented. Otherwise the existence of an Ombuds evaluation right may look more active than it is.

Complainants also need plain guidance on the difference between an Ombuds complaint and a Reconsideration Request. The former focuses on unfair treatment and voluntary resolution. The latter has defined eligibility, timing and decision consequences. Choosing the wrong route can matter, particularly because contact with the Ombuds is not formal notice of a claim.

Binding review answers a different question

The Independent Review Process exists to determine whether covered ICANN action or inaction violated the Articles or Bylaws. Current Article 4 describes it as final, binding arbitration to the extent allowed by law. It permits interim relief, including a stay, where specified standards are met, and decisions are intended to be enforceable in a court with jurisdiction.

Those features are exactly what the Ombuds lacks. An independent review panel can maintain the status quo, issue a binding determination and support court enforcement. The Ombuds can recommend that ICANN pause or reconsider, but cannot unilaterally preserve the claimant's position.

The comparison does not make independent review universally better. It is formal, legally complex and potentially expensive. Each party ordinarily bears its own legal costs, and standing, scope and procedural rules matter. Mediation can resolve a relationship in days that arbitration would harden over months.

The routes serve different needs. Use the Ombuds when the desired result depends on explanation, fair procedure, communication, voluntary correction or safe exploration. Use reconsideration when a qualifying Board or staff act should be reviewed under its defined standard. Use independent review when a covered violation requires a binding ruling or urgent protection. Use a court or contractual forum where applicable law or contract supplies the right.

These paths can interact, but they are not interchangeable. The Ombuds charter generally covers matters where independent review has not been invoked. The Framework also recognizes that pursuing a formal Article 4 or outside legal route may end the Ombuds matter. A person should not be pushed into confidential conciliation at the cost of missing an enforceable remedy.

ICANN should publish a remedy map organized by desired outcome. “I need an answer,” “I need a fair rehearing,” “I need this action paused,” “I need a binding interpretation,” and “I need compensation” should lead to different routes. Institutional acronyms are less useful to an affected person than remedy capability.

Soft accountability cannot bear a rights claim alone

Soft accountability works through persuasion, professional norms, publicity, relationships and the desire of institutions to appear fair. It is powerful where the respondent values legitimacy, the facts are uncertain, the harm is reversible and the parties expect to work together.

It is weak where interests are opposed and consequences are irreversible. A rejected application, delegated name, lost role, expired deadline or exclusion during a decisive vote may not be repairable through a later apology. A recommendation issued after the event can improve the future while leaving the complainant without relief.

Rights require more than sympathetic attention. They require a defined holder, duty-bearer, trigger, decision standard, deadline, remedy and enforcement route. An Ombuds finding can supply evidence that a duty was breached. It does not create the duty or enforcement mechanism by itself.

The office is also not legal counsel for complainants. Article 5 bars the Ombudsman from instituting, joining or supporting legal actions challenging ICANN. Neutrality would be compromised if the office became a party's representative. The result is that a person may receive route guidance without the resources needed to use the route.

This resource gap matters. A sophisticated applicant can distinguish an Ombuds complaint from reconsideration and independent review, preserve deadlines and retain counsel. A volunteer facing exclusion may not. Soft access can become unequal if only well-resourced entities can convert a fairness concern into enforceable protection.

The solution is not to turn the Ombuds into a court. It is to connect the office to rights without confusing roles: plain referral, deadline warnings, access assistance, translation, independent legal-information resources and a clear handoff record. Where the Ombuds identifies probable serious harm, the office should be able to flag urgency while the authorized forum decides interim relief.

Soft and hard accountability are complements. Confidential problem-solving can prevent unnecessary litigation. Enforceable review makes voluntary settlement credible because refusal has a possible consequence. Remove either side and the system degrades: every dispute becomes adversarial, or every remedy depends on goodwill.

What the office can actually correct

First, it can correct silence. A large share of perceived unfairness begins when a question, application or complaint receives no answer. The Ombuds can locate responsibility, request a response and set an expectation for timing. An answer may not satisfy the person, but it closes an information gap.

Second, it can correct factual misunderstanding. Access to records and interviews can reveal that a deadline was met, material was overlooked or an assumption was wrong. Parties may voluntarily revise their positions once they share the same facts.

Third, it can correct procedure. A selection, sanction or community decision can be reopened under the proper criteria, with notice and opportunity to respond. The Ombuds need not choose the winner to improve legitimacy.

Fourth, it can correct relational harm. Mediation, coaching and facilitated dialogue can produce apology, acknowledgement, communication rules and a workable future relationship. Courts rarely provide these outcomes.

Fifth, it can correct route confusion. Many submissions fall outside scope. A careful referral can prevent a person from spending months in the wrong forum, though the office must warn that contact does not preserve formal rights.

Sixth, it can correct institutional blindness. Aggregated cases reveal patterns in leadership selection, funding dependence, unanswered questions, incivility or exclusion. Annual and case reports can turn private experiences into governance evidence.

Seventh, it can increase the cost of refusal. A formal recommendation, response deadline and escalation to the Board create an audience and record. The office cannot compel ordinary compliance, but it can make noncompliance harder to conceal.

Eighth, it can support a formal mechanism without controlling it. The substantive evaluation of a Reconsideration Request can improve the committee's record and give the public an independent assessment.

These corrections are substantial. They concern daily legitimacy, not merely customer service. A multistakeholder body depends on volunteers believing that participation is meaningful, rules are applied consistently and conflict can be raised without retaliation.

But each correction depends on evidence of effect. Contact is not an answer. Mediation is not agreement. A finding is not remedy. A recommendation is not implementation. Referral is not access. Publication is not reform. The office should report each transition separately.

What it cannot safely be asked to replace

The Ombuds cannot replace an order staying imminent harm. Only a forum with that authority can preserve the status quo against a resistant decision-maker.

It cannot replace a binding determination of Bylaw compliance. Its fairness analysis may be relevant, but independent review has the legal mandate and enforceability.

It cannot replace contractual enforcement. A registry or registrar dispute may turn on agreement terms, compliance powers and specified dispute routes beyond the Ombuds remit.

It cannot replace employment procedure for ICANN staff. Personnel matters are excluded, and staff harassment reporting has distinct notice requirements.

It cannot replace Board responsibility. Even when advice is compelling, directors must decide whether to impose a sanction, change a policy or accept a recommendation.

It cannot replace public reasons where public power is exercised. Confidential settlement may resolve an individual's harm while leaving a policy question unanswered. Consequential Board action still needs a public rationale.

It cannot replace representation or counsel. Impartiality prevents the office from becoming the complainant's champion in litigation or negotiation over legal rights.

It cannot replace an implementation owner. Systemic observations matter only when an authorized body accepts responsibility, allocates resources and reports completion.

Finally, it cannot replace community accountability. Supporting organizations and advisory committees must maintain fair selection, participation and review practices of their own. Referring every conflict to one office would weaken responsibility at the place where conduct occurs.

A remedy ledger would make soft power auditable

ICANN should publish an anonymized Ombuds remedy ledger with stable categories. For each formal complaint or recommendation, it should record intake date, scope decision, route, time to first response, investigation status, outcome, remedy type, recipient response, implementation owner, target date, completion and follow-up.

The ledger should separate informal assistance from formal findings. Coaching, information, mediation, substantiated complaint and Board referral should not share one “resolved” label. Each demonstrates a different level of institutional intervention.

Recommendation status should include accepted, accepted with modification, rejected with reasons, no response, implementation under way, completed, effect verified and recurrence observed. A response deadline is meaningful only if missed deadlines are counted.

Confidentiality should determine granularity, not erase the row. Highly sensitive cases can be grouped by quarter or remedy class. Small categories can be suppressed under a published rule. An independent reviewer can test the underlying record and certify aggregate accuracy.

The ledger should report power boundaries. If the requested remedy was unavailable, state whether the complainant received a referral and deadline warning. If a person chose a self-managed route, offer an optional follow-up and record whether outcome is unknown.

Board interaction needs special visibility. Record own-initiative investigation requests, authorization or refusal, serious-fairness notifications, recommendation responses, harassment referrals, recusals and budget decisions. These are the moments when independence is tested.

User experience should be measured separately from outcome. Respectful treatment, clear communication and timeliness are service qualities. Correction, implementation and nonrecurrence are remedy qualities. A complainant may be treated well and still lose; a recommendation may be accepted despite a poor experience.

External evaluation should sample both successful and failed matters. Success-only case studies inflate effectiveness. Refusals, withdrawals, route errors, late responses and recurring conduct reveal where the office's design needs support from enforceable mechanisms.

Reform should strengthen consequence without turning the office into a tribunal

The first reform is full implementation of the 2026 Framework. Published response times, case categories and outcome definitions should appear in annual measures, and the Work Stream 2 status page should be reconciled with the Board-approved document.

Second, ICANN should complete an independent Ombuds Advisory Panel for appointment advice, performance feedback and periodic evaluation, while barring panel involvement in individual cases. This would diversify oversight without creating a second Ombuds.

Third, formal recommendations should receive a public substantive response by default. Confidentiality may require redaction or aggregation, but acceptance, rejection, delay and implementation should be visible.

Fourth, own-initiative investigations within clearly defined jurisdiction should not depend entirely on prior permission from a body that may be affected by the inquiry. At minimum, any refusal should be reasoned, published and reviewable by an independent authority.

Fifth, referral guidance should be remedy-based and carry conspicuous deadline warnings. Every acknowledgement should state that Ombuds contact is not formal notice and does not itself provide a stay.

Sixth, ICANN should fund accessible legal-information and claimant-support resources separate from the Ombuds. The office can remain neutral while people receive help understanding enforceable routes.

Seventh, the Board should define how it handles Ombuds findings in harassment and serious-fairness matters: recusal, evidence review, response time, available actions, confidentiality and publication. Discretion without a visible method can undo the office's work.

Eighth, annual reports should follow systemic suggestions through later years. Leadership selection, feedback loops and participation barriers identified in FY25 should return with adoption and effect evidence, not disappear when the next theme is chosen.

Ninth, the office should publish budget requests and final allocations in a comparable series. Independence requires resources, and resource dependence is observable.

Finally, claims about accountability should remain modest. The Ombuds is a valuable resolver, investigator and institutional sensor. It is not a guarantee that ICANN action can be stopped or reversed. Accurate language preserves trust when a hard case reaches the boundary.

Conclusion

An ombudsman without order power can correct more than the title suggests. The office can obtain information, expose a broken procedure, help a group reconsider, mediate a durable agreement, validate or reject a complaint, recommend repair, reveal a systemic pattern and force an institution to answer publicly.

Those effects are particularly valuable in ICANN, where much governance depends on repeat relationships among volunteers, staff, directors and organized communities. A low-cost, confidential route can rescue participation before conflict becomes litigation or departure.

The office cannot compel the ordinary remedy on its own. It cannot set aside a Board act, grant interim protection, award damages or convert a recommendation into a binding right. Even when it identifies harassment, imposed action ultimately comes from the Board. When a covered Bylaw violation requires a final ruling, independent review supplies powers the Ombuds does not.

The 2026 Framework improves role clarity and response expectations. FY25 evidence shows growing use, heavy demand outside scope, a strong collaborative character and substantial uncertainty about independence. The next advance must be evidence of uptake: who answered, what changed, when it changed and whether the harm recurred.

Soft accountability is not failed adjudication. It is a different instrument. It succeeds when it makes voluntary correction faster, fairer and more likely, and when it hands unresolved rights claims to a forum that can bind. It fails when its availability is used to imply that enforceable relief exists where none does.

ICANN should protect the Ombuds precisely by refusing to make the office carry a burden it was not designed to bear. Fair listening, skilled mediation and reasoned recommendation are essential. Rights, stays and binding remedies remain essential too.

Sources