Summary

  • ICANN's Independent Review Process is stronger than public consultation because it can test conduct against the Articles and Bylaws through an adversarial proceeding before neutral decision-makers and produce a reasoned public declaration.
  • Independence is real but bounded: standing, excluded subject areas, party-selected panelists, legal expense, language, procedure and years of delay determine who can obtain a merits ruling.
  • Enforceability is divided between a binding determination that ICANN violated its governing commitments and a narrower remedial power that often stops short of granting an application, cancelling a contract or substituting the panel's preferred policy result.
  • The decisive accountability question is therefore not whether ICANN sometimes loses, but whether an eligible claimant can afford to reach a timely decision and whether the institution's response restores the choice that the violation distorted.

A constitutional hearing, not another listening exercise

The Independent Review Process, usually shortened to IRP, occupies an unusual place in global Internet governance. ICANN is a California nonprofit corporation with worldwide influence over the domain-name system. It makes decisions through a mix of supporting organisations, advisory committees, public comments, staff work, contractual negotiations and Board resolutions. Most of those channels are participatory rather than adjudicative. They give affected groups opportunities to speak, but they do not ordinarily place ICANN before a neutral tribunal that must decide whether its conduct complied with a controlling text.

The IRP does. Under Article 4, Section 4.3 of the ICANN Bylaws, an eligible claimant may challenge covered actions or failures to act that violate ICANN's Articles of Incorporation or Bylaws. The current design promises a meaningful, affordable and accessible form of expert review; consistent and coherent interpretation of ICANN's constitutional commitments; publicly available decisions that can guide later conduct; and final resolutions capable of binding the parties and being enforced in court. Those are legal and institutional ambitions, not the vocabulary of a conventional consultation.

That difference matters. A public comment entity supplies an opinion to the decision-maker. An IRP claimant alleges a violation by the decision-maker. The claimant can present evidence and argument, confront ICANN's account of events, seek interim protection in defined circumstances and receive a reasoned declaration from people who are not members of the Board whose conduct is under review. A consultation can expose disagreement. An IRP can establish noncompliance.

But the distinction should not be romanticised. Arbitral review is not broad public accountability in miniature. It is a specialised cause of action with jurisdictional limits, procedural burdens and costly factual questions. It can say that ICANN acted inconsistently with its governing commitments while leaving the institution substantial discretion over the next lawful step. Its strength lies in disciplined review and an enforceable legality judgment. Its weakness lies in access, duration and the distance between a declaration of violation and restoration of the opportunity that was lost.

The fairest assessment separates three questions that are too often collapsed. Is the decision-maker independent enough to judge ICANN? Can an affected person or organisation afford to obtain judgment? Does a successful ruling compel a practical correction? ICANN can perform well on one axis and poorly on another. An independent panel does not make lawyers inexpensive. Reimbursement of administrative fees does not return years to an applicant. A binding finding does not necessarily transfer a contested top-level domain or reopen a signed registry agreement.

What independence means in the current rules

The post-2016 Bylaws give the IRP a more judicial character than its earlier recommendation-centred form. The standard of review is objective and de novo: the panel does not simply ask whether the Board had some basis for its action. It determines whether covered conduct violated the Articles or Bylaws. At the same time, the Bylaws preserve a boundary around reasonable Board business judgment. The panel is not authorised to replace a reasonable judgment with its own merely because it would have chosen differently.

Institutional design supports that separation. The Bylaws call for a standing panel of at least seven members with relevant legal and subject expertise. Parties select panelists through the prescribed method rather than accepting a tribunal appointed solely by ICANN. The dispute-resolution provider must be independent of ICANN, and panel decisions are to be reasoned and generally public, subject to protections for confidential material. In September 2024, ICANN announced the first 12 members of the standing panel, after a selection effort that drew 99 expressions of interest and a Board-approved slate. Establishing that bench addressed a long-running gap between the reformed constitutional design and day-to-day operation.

Independence, however, is not the same as complete institutional separation. ICANN helped design the Bylaws, participates in the community group that refines procedural rules, contracts with the dispute-resolution provider and pays the tribunal's administrative costs. That is not proof that panelists lack neutrality. It does mean that independence rests on appointment safeguards, professional duties, transparent reasoning and the parties' ability to challenge conflicts rather than on the existence of a separate public court system.

Party selection can strengthen confidence because neither side controls the full panel. It can also reproduce features of commercial arbitration: specialist lawyers, extensive submissions, disputes about confidentiality and procedure, and a market in experienced counsel. A well-resourced registry applicant may regard those features as familiar. A community group or person affected by a governance decision may see a daunting threshold. Formal equality between parties is not the same as equal capacity to investigate facts, retain experts and sustain years of advocacy.

The written record nevertheless changes ICANN's incentives. Board minutes and consultation summaries that might be sufficient for ordinary political explanation can be tested against duties of transparency, reasoned decision-making, fairness and faithful adherence to established procedures. A panel can distinguish a conclusion from the evidence said to support it. It can ask who actually made the decision, whether the proper body exercised the authority and whether relevant considerations were faced. Even when the requested substantive result is denied, that examination creates a public account that consultation alone rarely supplies.

Standing turns public concern into a private threshold

The Bylaws do not allow anyone who dislikes an ICANN decision to bring an IRP. A claimant must be a natural or legal person, group or entity materially affected by a covered action or failure to act. The alleged injury or harm must be direct and causally connected to the challenged conduct. The Empowered Community is deemed materially affected in the circumstances specified for it, but most claimants must prove their own relationship to the dispute.

That requirement protects the IRP from becoming a general forum for abstract objections. It also narrows the accountability population. Many ICANN decisions produce dispersed effects. A change to registry contract terms may alter incentives across registrants, registrars and future users without imposing a large, immediate loss on any one person. A failure to follow a consultation promise may diminish trust across the community while leaving no entity able to quantify individual harm. The more collective the interest, the harder it may be to convert the concern into adjudicative standing.

The final declaration in Namecheap v ICANN is instructive. The panel did not accept that an abstract interest in lawful governance was enough. It did, however, recognise that procedural injury can support standing where the claimant has a concrete interest and presents a reasonably credible account of causal harm. Namecheap participated in the affected market as a registrar and paid registry fees. That economic and procedural connection was materially different from speaking only as a member of the general public.

This is an important middle position. If standing required proof that the claimant would certainly have obtained a different result, review of defective procedure would become almost impossible: the missing lawful procedure is precisely what makes the alternative result unknowable. If any concerned observer could claim injury, the IRP could become an open-ended constitutional complaint desk. The Namecheap approach allows credible procedural harm connected to a real interest without treating public dissatisfaction alone as sufficient.

The cost of that line is underrepresentation of diffuse interests. Registrants are numerous and individually small. Future applicants may be affected before their interests crystallise. Technical users can depend on institutional reliability without holding a contract that makes injury easy to show. Supporting organisations and advisory committees can sometimes organise collective positions, but an IRP is not a class action. Its standing rule asks who was harmed, not simply whether ICANN's constitutional discipline matters to everyone.

Standing decisions therefore deserve as much attention as merits awards. A strong declaration in a well-funded applicant dispute demonstrates that the tribunal can resist ICANN. It does not establish that less concentrated harms are reviewable in practice. A credible accountability assessment must count claims dismissed or narrowed before the merits and must ask what kinds of affected communities never filed because standing was uncertain.

Jurisdiction has deliberate blind spots

The IRP is also limited by subject matter. The Bylaws exclude several categories, including claims concerning country-code top-level domain delegations and redelegations, Internet numbering resources, protocol parameters and decisions on ICANN's grant-making programme. Certain challenges involving consensus-policy development require support from the relevant supporting organisation. Those boundaries reflect the distribution of authority across the wider Internet governance environment, but they prevent the IRP from serving as a universal tribunal for every ICANN-associated controversy.

The exclusions are particularly important when public descriptions call the IRP ICANN's ultimate accountability mechanism. Ultimate within its covered field is not comprehensive across all fields. Number-resource disputes have distinct institutions and policy communities. Country-code delegation questions involve governments, local communities and long-established practices. Protocol parameters connect to technical stewardship arrangements. The Bylaws make institutional choices about where review belongs; they do not convert every contested exercise of coordination into an IRP claim.

Even within jurisdiction, the panel reviews compliance rather than wisdom at large. A policy may be unpopular, commercially harmful or technically questionable without violating a specific constitutional commitment. Conversely, ICANN may pursue an outcome within its Mission and still violate the Bylaws by using the wrong decision-maker, disregarding an established procedure or failing to act transparently. The tribunal's legitimacy depends on maintaining that distinction. If every policy disagreement became a constitutional violation, adjudicators would displace community policy formation.

If procedural commitments were treated as optional whenever the Board preferred the result, accountability would become ceremonial.

This bounded jurisdiction explains why arbitral review can be both powerful and incomplete. It is powerful where a claimant identifies a covered duty and proves a concrete violation. It is incomplete where a grievance concerns excluded resources, diffuse public interest or the quality of a judgment that remained within lawful discretion. Public accountability must therefore include other institutions: transparent Board records, meaningful consultation, empowered-community powers, reconsideration, ombuds functions, contractual compliance and the courts where applicable. The IRP is a central column, not the entire building.

Affordability is more than the tribunal invoice

ICANN bears the administrative costs of maintaining the IRP, including the standing panel. For an ordinary case, each side generally pays its own legal expenses, while ICANN ordinarily pays provider and panel costs. A panel may shift costs when a claim or defence is frivolous or abusive. Community IRPs receive special treatment under the Bylaws. The rules also state that ICANN should seek reasonable access for nonprofit and community claimants.

Those commitments make the mechanism more accessible than an arrangement in which a claimant must finance both its lawyers and half of an expensive tribunal from beginning to end. Published awards show that provider and panel charges can reach hundreds of thousands of dollars. ICANN's assumption or reimbursement of those charges is meaningful. Yet legal fees, experts, document review, executive time and opportunity cost can exceed the filing and arbitrator bill. A rule that says each side bears its own lawyers leaves the largest access barrier largely where it began.

The monetary figures in major cases illustrate scale without supplying a complete affordability measure. In the .AFRICA dispute brought by DotConnectAfrica, the panel recorded administrative fees of $4,600 and panel fees and expenses of $403,467.08. ICANN was directed to reimburse DotConnectAfrica $198,046.04 for its share, while each party retained its own legal expenses. In Amazon's challenge concerning its applied-for string, ICANN was directed to reimburse $163,045.51 in proceeding costs.

In the first Afilias .WEB dispute, ICANN ultimately faced reimbursement obligations that included $479,458.27 for Afilias's share of IRP costs and $450,000 associated with legal fees for emergency relief. Those numbers show that cost allocation can prevent tribunal charges from making a successful challenge hollow. They also show why only counting the initial filing fee would misdescribe the undertaking.

Namecheap's 2022 declaration reported panel costs of $841,895.76 and administrative fees of $13,835. ICANN was required to reimburse Namecheap's $58,750 share of specified administrative charges, while each side generally remained responsible for its own legal representation. A nearly three-year case about registry agreements involved extensive legal and economic material. The published allocation reveals the cost of adjudication, but not the claimant's full expenditure or the burden on people who chose not to file.

No publicly established denominator shows how many eligible parties considered an IRP and abandoned it because of cost. Nor is there a consistent public series for claimant legal spending, financing arrangements or staff time. It is therefore not possible to calculate a defensible affordability rate from the awards alone. Successful corporate claimants prove that the mechanism is usable by sophisticated parties. They do not prove that it is meaningfully affordable for small nonprofits, registrant groups or individuals.

Affordability also includes downside uncertainty. The Interim Supplementary Procedures permit summary dismissal where standing is absent and provide standards for interim measures. A claimant must assess whether a panel will hear the merits, whether urgent relief is available and whether its position could be labelled frivolous or abusive. The Bylaws also connect cost consequences to good-faith participation in the voluntary Cooperative Engagement Process in specified circumstances. Each threshold can be defensible on its own while adding legal advice before the central issue is ever heard.

An access policy should consequently be measured against total participation cost, not ICANN's share of arbitrator fees. Useful public indicators would include time to panel formation, claimant legal-cost ranges reported voluntarily, use of fee support, outcomes for nonprofit claimants, translation needs, summary dismissals and the point at which cases settle or are withdrawn. Without those data, the constitutional word "affordable" remains an objective supported by some cost protections rather than a demonstrated condition for the full affected community.

Delay can decide the dispute before the panel does

The Bylaws state that an IRP should ordinarily produce a written decision within six months after the panel is constituted, unless fairness requires more time. They also make clear that failure to meet that target is not itself a basis for another challenge. The aspiration recognises a basic truth: a lawful answer delivered after the relevant commercial or governance moment may have far less remedial value.

Actual major proceedings routinely last longer. DotConnectAfrica commenced its IRP in October 2013 and received a final declaration in July 2015. Amazon initiated its case in March 2016 and received a final declaration in July 2017. Afilias began its .WEB proceeding in November 2018; the merits hearing took place in August 2020, and the principal declaration arrived in 2021, followed by corrections and further decisions. Namecheap filed in February 2020 and received the final declaration in December 2022.

The continuing Altanovo .WEB matter demonstrates the modern problem even after procedural reform. ICANN's IRP and Cooperative Engagement status report dated 25 June 2026 records receipt of the request in July 2023, constitution of the full panel in March 2024 and a merits hearing in November 2025, with the case still active in June 2026. The sequence may reflect complexity, extensive evidence and the parties' procedural choices. It nevertheless places the six-month aspiration far from the experienced calendar.

Time has distributive effects. A registry applicant can lose financing, staff or market position while a string remains unresolved. A challenged contract may be signed and performed. Community volunteers rotate out of leadership roles. Evidence becomes harder to reconstruct, while ICANN staff and Board membership change. Even a prevailing claimant may return to a decision environment transformed by years of reliance and sunk costs.

Interim measures are intended to prevent irreparable harm where the defined test is met. They are essential but cannot solve every timing problem. A broad freeze can impose costs on other applicants, registries and users, so panels must weigh competing harms. Denying a freeze may allow the contested arrangement to harden. Granting one may preserve review but delay an entire programme. The difficulty is not merely procedural speed; it is that the DNS continues to operate while legality is contested.

Delay also affects bargaining power. A claimant spending money each month may settle for a narrower accommodation. ICANN may face pressure to preserve arrangements that other parties now rely upon. Public attention moves elsewhere. None of those effects proves bias in the tribunal. They show that adjudicative independence cannot compensate indefinitely for elapsed time. A mechanism should be judged not only by whether the final declaration is careful, but by whether a practical field of choice still exists when it arrives.

DotConnectAfrica: a clear violation, followed by renewed evaluation

The DotConnectAfrica case remains a foundational demonstration that the IRP can make ICANN listen. The dispute concerned DotConnectAfrica's application for .AFRICA and the effect of Governmental Advisory Committee advice. In its July 2015 declaration, the panel concluded that Board actions and failures to act were inconsistent with ICANN's Articles and Bylaws. It recommended that ICANN refrain from further action on the competing application and permit DotConnectAfrica's application to resume evaluation.

On 16 July 2015, the Board accepted the findings and directed the evaluation to resume. The result was not an award of .AFRICA to the claimant. It restored a process from which the application had been removed. That distinction captures both the potency and limit of the mechanism. The panel identified a constitutional failure and changed the next institutional step. It did not itself make the technical and policy determinations necessary to delegate a top-level domain.

The case strengthened accountability in at least three ways. First, it showed that deference to high-level governmental advice did not eliminate the Board's duties. Second, the public declaration reconstructed decision-making that ordinary records had not adequately explained. Third, ICANN bore the proceeding costs allocated by the panel and publicly recorded its response. Those consequences exceed what a comment submission could achieve.

Yet restoration to evaluation is not restoration of time. The claimant had already spent almost two years in review, and the broader contest over .AFRICA continued in other legal venues. The substantive entitlement still depended on later evaluation and litigation. The IRP corrected an institutional wrong without resolving every commercial and legal consequence caused by it. That is a recurring pattern: a strong legality finding, an important procedural reset and an incomplete repair of the claimant's position.

Amazon: independent evaluation as a constitutional requirement

Amazon's challenge concerned the Board's treatment of Governmental Advisory Committee consensus advice opposing Amazon's applications for the .AMAZON strings. A majority of the panel found in July 2017 that Amazon prevailed. The declaration focused on the Board's failure to exercise the independent and objective judgment required by ICANN's governing commitments rather than merely accepting the advice as conclusive. ICANN was directed to reimburse Amazon $163,045.51 in specified proceeding costs.

The award did not require immediate delegation of the strings. It required the Board to reconsider the applications by exercising its own judgment. That outcome is constitutionally significant. Advice can carry special weight without erasing the duty of the legally responsible body to understand reasons, assess compatibility with ICANN's Mission and explain its decision. The panel enforced a distinction between listening to governments and surrendering the decision.

The later history was long and contested, involving further Board consideration, proposals and engagement among Amazon and the affected governments. That complexity does not negate the award. It demonstrates that a declaration can reopen a decision without controlling the political settlement that follows. For supporters of review, that is appropriate institutional restraint. For a claimant seeking a definite commercial entitlement, it can feel like winning the rule and returning to the same arena.

Amazon therefore helps define enforceability accurately. The obligation to undertake lawful independent evaluation was not optional advice. The panel's conclusion carried constitutional force. But the choice produced by that evaluation remained the Board's, provided the Board complied with its duties. Enforceability attached most strongly to the standard and the need to decide again, not to the claimant's preferred result.

Afilias and .WEB: the remedy line drawn in public

The first Afilias .WEB IRP exposed the same separation in a more expensive and procedurally intense form. Afilias challenged ICANN's handling of Nu Dot Co's application and alleged undisclosed control associated with Verisign. After an extended hearing, the panel found serious failures in ICANN's treatment of the matter. Its corrected final declaration concluded that ICANN had violated its governing commitments in relevant respects.

Afilias sought a decisive remedy: disqualification of Nu Dot Co and an outcome that would move Afilias toward the string. The panel declined to order that result. It reasoned that ICANN had to make the first substantive determination through a lawful process. Again, adjudicators could identify noncompliance and direct attention to the decision that had not been properly made, but they would not take over the programme administrator's role.

The cost decisions were substantial. ICANN was required to reimburse $479,458.27 representing Afilias's share of IRP costs and $450,000 in legal fees associated with emergency relief. Later, a separate decision found a correction request frivolous and required Afilias to pay $236,884.39 in ICANN legal fees. Together, the rulings show that cost allocation cuts both ways. It can make a successful challenge more meaningful, and it can penalise procedural conduct that the panel regards as abusive.

The case also demonstrates why legal expense cannot be treated as a peripheral concern. Only parties able to litigate complex questions of ownership, disclosure, auction conduct and programme rules could sustain such a dispute. The public benefited from an exacting examination of ICANN's responsibilities. Access to that examination depended on a corporate claimant with enough at stake to finance it.

For institutional legitimacy, the crucial point is not that the panel refused Afilias's desired substantive award. A tribunal that automatically granted the claimant's preferred programme result whenever it found a procedural violation would risk exceeding its constitutional role. The concern is whether the subsequent ICANN decision is timely, transparent and genuinely open. If the same outcome returns after lawful consideration, the institution should be able to show what changed in the reasoning even if the result did not change.

Namecheap: a victory that did not restore price controls

Namecheap's challenge to the 2019 renewals of the .ORG and .INFO registry agreements is the clearest modern test of the gap between declaration and remedy. ICANN staff negotiated and executed agreements that removed historic price-control provisions despite overwhelming opposition in public comments. Namecheap argued that ICANN failed to act openly and transparently, allowed staff to make a policy choice that belonged to the Board and disregarded procedures required by its governing commitments.

The panel agreed with important parts of that case. It found that approval of the agreements without the previous price caps was inconsistent with ICANN's Articles and Bylaws in several respects. The decision had not been made in the open and transparent manner required, and the removal of price controls involved a policy judgment that should have received proper Board treatment. The declaration is powerful because it did not merely request that ICANN take comments more seriously. It adjudicated who had authority, what procedures applied and why the conduct was unlawful.

The panel also stated that it lacked authority to annul the executed agreements or impose the remedy Namecheap sought. Ten-year contracts had been signed. The tribunal could declare the violation and recommend responsive action, but it would not rewrite the registry agreements itself. That line protected the division between constitutional review and contract administration. It also meant that the commercial effect at the centre of the dispute remained in place.

The Board's response unfolded over almost two more years. In June 2023 it commissioned economic analysis. In November 2024, after receiving the resulting work and further material, it decided not to pursue restoration of price controls. The Board directed attention to possible improvements in clarity around decision-making and public comment, but the key contract terms survived.

Namecheap thus won an authoritative judgment about institutional conduct and lost the practical result it most wanted. That is neither a meaningless victory nor full correction. The declaration establishes precedent, disciplines future allocations of authority and gives the community a detailed account of failure. It may make later staff and Board action more careful. But registrants did not receive renewed price caps, and the affected contracts were not unwound.

The case is a warning against binary measures of compliance. If compliance means that the Board considered the declaration, commissioned analysis and issued reasons, ICANN complied. If compliance means that the institutional position was returned to the moment before the unlawful agreements were signed, it did not. Both statements can be true because the Bylaws bind the legality determination more tightly than they prescribe the substantive cure.

What exactly is binding?

The current Bylaws state that IRP declarations are binding final decisions to the extent permitted by law and are intended to be enforceable in court. If the Board rejects or declines to follow a declaration, it must explain itself, and a prevailing claimant may seek judicial enforcement. This court backstop is a major improvement over a model in which the panel merely offers advice that the Board can ignore without legal consequence.

The word "binding" still requires careful parsing. A panel may declare that ICANN violated its Articles or Bylaws. It may recommend that a challenged action be stayed or that the Board take corrective steps. It may allocate costs. But it cannot always award the underlying top-level domain, cancel a third-party contract, substitute its economic policy or order action beyond its authority. A court can enforce what the declaration lawfully decides; it cannot enlarge the declaration into a remedy the panel never had power to grant.

This produces a layered structure. At the first layer, the constitutional finding is authoritative. At the second, ICANN must consider and respond to the declaration in a transparent Board act. At the third, the institution chooses among lawful substantive options unless the declaration specifies a narrower mandatory step. At the fourth, a court can be asked to enforce the binding obligation if ICANN refuses. Each layer adds accountability, but each also consumes time and money.

Third-party rights complicate enforcement further. A registry agreement may affect an operator that was not the sole target of the claim. A string dispute may involve competing applicants or governmental interests. Undoing a completed transaction can harm parties that relied on it. Panels therefore tend to protect the distinction between deciding whether ICANN complied and administering the programme in ICANN's place. That restraint is institutionally sensible, yet it can make the prevailing claimant's remedy depend on the same Board whose prior conduct failed review.

The quality of Board compliance should be judged by more than formal acknowledgment. A serious response identifies every material finding, states the available remedies, discloses constraints, compares alternatives, explains the chosen action and sets a timetable. It should say whether the decision was returned to an open state or whether reliance interests foreclosed options. Where the Board retains the same result, it should distinguish a lawful reconsidered outcome from simple defence of the past.

Consultation, reconsideration and IRP are not substitutes

Public comment is broad, cheap and open. It can gather technical expertise and expose distributional effects before a decision. Its weakness is that ICANN remains the listener, interpreter and decision-maker. There is ordinarily no neutral finding that the response summary misunderstood a submission or that the chosen authority lacked power. Participation can improve judgment, but it does not guarantee an adjudicated answer.

Reconsideration is more focused. It asks ICANN to revisit a staff or Board action on defined grounds and can produce a Board Accountability Mechanisms Committee recommendation. It is generally faster and less expensive than full arbitration. Yet it remains an ICANN decision about ICANN conduct. The Board may correct an error, but the mechanism does not offer the same distance as a neutral panel.

The IRP is narrower and heavier. It requires standing, a covered constitutional claim and sustained advocacy. In exchange, it produces independent findings, reasoned precedent and an enforcement path. Its value is greatest where the dispute is not simply about persuading the institution but about proving that the institution crossed a legal boundary.

These mechanisms should form a sequence of complementary protections. Consultation should prevent error while choices remain open. Reconsideration should correct identifiable mistakes quickly. IRP should adjudicate serious constitutional failures that survive or evade those stages. Courts should be a last backstop when a binding declaration is not honoured. If early mechanisms are weak, claimants are pushed toward the most expensive one. If the IRP is slow, earlier unlawful commitments harden and make remedy more difficult.

The availability of IRP should never excuse poor consultation. Telling affected users that they can arbitrate after a procedurally defective contract is signed reverses the economics of accountability. ICANN controls the initial decision calendar and record; a claimant bears the burden of challenging it. The more the institution invests in open reasons and responsive participation before action, the less often legitimacy depends on a multimillion-dollar dispute years later.

How to measure compliance without declaring victory too soon

ICANN's IRP document page provides pleadings, orders, declarations and status materials for many cases. That transparency allows observers to reconstruct claims and outcomes better than would be possible in confidential commercial arbitration. Yet the public record is not a ready-made accountability scorecard.

A useful measure starts with access. How many requests are filed, how many reach the merits, how many are dismissed for standing or timeliness, and how many are withdrawn? Which claimants are companies, community bodies, nonprofits or individuals? How long does panel appointment take? Are translations and accessibility measures used? Published merits victories alone select for parties that crossed every prior barrier.

The second measure is elapsed time. Dates should be reported from the request, panel constitution, interim-relief decision, hearing, final declaration, Board response and completed remedial action. A six-month decision after a two-year panel-selection dispute is not a six-month remedy. Nor is a fast declaration enough if the Board takes years to decide what it will do.

The third is total cost. Provider and panel charges are visible in many declarations, but legal spending and organisational burden are often not. Voluntary reporting in ranges could protect confidential strategy while showing whether access is improving. Any fee-support arrangement should disclose eligibility, uptake and outcomes without exposing private claimant information.

The fourth is remedial restoration. Did the claimant obtain only a declaration, a renewed evaluation, an interim stay, reimbursement, a changed policy, a modified contract or the substantive result sought? Did third-party reliance make full restoration impossible? Did ICANN repeat the same outcome with better reasons, or did it materially reopen the choice? These categories avoid treating every claimant victory as equal.

The fifth is precedent. Later Board papers and staff decisions should cite relevant declarations and explain their application. A finding that changes future conduct can have value beyond the immediate remedy. Conversely, a celebrated award that disappears from later reasoning has little systemic force. Precedent need not make policy immutable; it should make constitutional interpretation coherent.

Reform should target access, speed and remedial clarity

The basic independence architecture should be preserved: qualified panelists, conflict safeguards, party participation in selection, de novo constitutional review, public reasons and a judicial backstop. Reforms that make the Board the final unreviewed interpreter of its own compliance would sacrifice the mechanism's distinctive value.

Access requires more than lower filing charges. ICANN and the community should define practical support for eligible nonprofit and community claimants, including early jurisdiction advice, transparent fee assistance, translation, accessible hearings and proportionate procedure. Small claims should not require the same documentary scale as disputes over contested registry applications. Proportionality can reduce cost without lowering the standard of fairness.

Speed requires active case control. Early conferences can isolate standing and jurisdiction, narrow document disputes and set a realistic merits calendar. Panels should explain departures from the six-month target in public scheduling orders. Repeated extensions may be justified, but their causes should be visible: party requests, panel availability, settlement talks, evidence complexity or provider delay. That information would show where reform can work.

Remedial clarity is equally important. Claimants should know which forms of relief are legally available before spending years litigating. Declarations should separate binding findings, mandatory acts, recommendations and matters reserved to the Board. ICANN's response should use the same categories. Where a contract or delegation makes restoration difficult, the Board should explain when reliance arose and why interim protection did or did not preserve alternatives.

The institution should also avoid creating irreversible facts while a serious, timely claim is pending. That does not mean every request deserves a freeze. It means the decision calendar should recognise the possibility of review, and panels should be able to address urgent preservation questions quickly. A remedy is most effective when it protects choice before sunk costs make correction disproportionate.

Finally, compliance should be independently reviewable as a distinct phase. A short, defined procedure could resolve whether ICANN carried out a mandatory declaration without reopening the entire merits dispute. Court enforcement remains vital, but requiring a claimant to begin another costly judicial contest for every ambiguous response weakens the benefit of winning the IRP.

The institutional verdict

ICANN's Independent Review Process is genuinely independent in the sense that matters most: it can place the institution's conduct before neutral adjudicators, test that conduct against superior governing commitments and publish a reasoned finding that the Board cannot treat as an ordinary comment. DotConnectAfrica, Amazon, Afilias and Namecheap all show that ICANN can lose on consequential questions. Cost awards and the court-enforcement language give those findings weight.

It is only partly affordable. ICANN's payment of tribunal charges and reimbursement orders reduce a major barrier, but the party-funded legal and expert burden remains large. The public record does not show that ordinary affected communities can reliably carry that burden. Standing and jurisdiction further select which harms can be heard.

It is enforceable most strongly at the level of constitutional legality and less consistently at the level of substantive restoration. A binding finding can require ICANN to reconsider, explain or perform a defined act. It often cannot give the claimant the string, annul the agreement or select the policy outcome. By the time the Board responds, delay and third-party reliance may have narrowed the lawful alternatives.

That combination still makes IRP stronger than consultation. Consultation asks ICANN to listen. IRP can establish that ICANN failed its own rules, assign financial consequences, create precedent and support resort to a court. But broad public accountability asks an additional question: can people without concentrated commercial stakes obtain a timely remedy before the disputed action becomes irreversible?

The answer remains uneven. The IRP has matured from a recommendation-oriented safeguard into a constitutional adjudication mechanism with real authority. Its next test is not whether panels can write forceful declarations. They can. It is whether standing, cost, delay and compliance practice allow those declarations to protect the full range of interests that ICANN's legitimacy depends upon.

Sources