Summary

  • The IAB’s 15 October 2026 liaison asks ICANN’s Board to condition the proposed model’s initial phase on a binding review panel outside the Council for operator designation, removal and performance suspension.
  • ICANN’s February Functional Model records independent review as a future consideration for the Establishment Phase. That is a documented sequencing choice, not proof that the process is settled or that the community agrees with the IAB.

A governance council can be formally created before anyone knows how to challenge its most consequential decisions. That is the sequencing risk at the centre of a 15 October intervention by the Internet Architecture Board (IAB) on the proposed Root Server System Governance Framework. The IAB asks the ICANN Board not to move into the model’s first phase until core accountability safeguards—including an independent appeal route—are addressed. Its letter is advice to the Board, not a binding decision or a declaration of community consensus.

The underlying framework is deliberately staged. ICANN’s February 2026 Functional Model says the Initiation Phase forms the Council and Secretariat, creates a limited forum for organization and preparation, and does not yet exercise the model’s full governance powers. The Establishment Phase is meant to formalize functions, policies and accountability mechanisms. Only after milestones are met would the Governance Phase give the structure authority over RSS governance. The authors present staging as a way to preserve operational continuity while governance capacity grows; it is not simply a plan to grant the Council immediate control.

The narrow problem is what happens if the first phase creates an institution whose composition, authority and working habits become costly to alter before an external remedy exists. The Functional Model describes appeals and review mechanisms as internal to the Council in the relevant draft. In its disposition of consultation feedback, it marks independent review “for future consideration” during Establishment. The IAB’s requested change is more specific: a neutral external panel, binding within its remit, available before Initiation for challenges to operator designation, removal and performance suspension.

An internal appeal and an independent review are not interchangeable. The former asks the institution to reconsider itself; the latter creates a route outside that decision-making body.

This matters because the proposed Council would sit at the intersection of technical continuity and public accountability. The framework gives direct Council participation to Root Server Operators (RSOs), the country-code and generic top-level domain registry communities, with other interested groups able to engage through consultation and liaisons. It also recognizes that RSOs perform the service and that technical expertise is essential. The IAB argues that the draft gives RSOs half the voting power, permits potential overlap across groups, and leaves some affected communities without formal representation.

Those are the IAB’s criticisms of the model, not uncontested findings about capture.

The appeal question cannot be separated from the powers an appeal can reach. The framework assigns later functions for designation and removal, performance monitoring, finance and resource management, security reporting, and policy advice. During Establishment, milestones call for due process and appeals to be defined; a later transition requires active functions, ratified core policies, a public assessment and consultation. That creates a real safeguard sequence before the Council reaches its fullest authority.

But it does not answer whether a person or operator affected during Initiation can challenge the initial rules, interim decisions, or a Council interpretation before the independent mechanism exists.

The IAB also wants a binding diversified funding plan before the Council is chartered, stronger checks on the Council, more balanced voting, and no double voting by organizations represented in multiple groups. These recommendations should be evaluated together because remedies depend on who appoints reviewers, who funds them, what decisions are reviewable, and whether their rulings bind the institution. A panel that exists on paper but has no protected budget, published reasons or jurisdiction over consequential decisions would not solve the credibility problem.

Conversely, imposing a fully empowered review system before operational roles are clear could create delay or confusion. The design challenge is to define a narrow, usable appeal route without making routine technical operations hostage to procedural disputes.

There is also a distinction between setting up a Council and authorizing it to govern. The February model says Initiation is limited in scope and authority; later governance depends on completing milestones and consulting stakeholders. That distinction supports the case for staging, but it does not settle whether postponing the independent panel is safe. Institutions build precedent while they are still “preparatory”: early choices shape expectations, budgets, appointments and what later participants regard as changeable.

The more those choices define access to decision-making, the greater the burden to show a real amendment route rather than a promise to revisit them.

For the Board, the choice is not simply “start” or “stop”. It can authorize organizational preparation while making the next phase conditional on a published appeal charter: independent appointments, stable funding, a defined scope, standing for affected operators or communities, reasoned decisions, deadlines, disclosure rules and a remedy the Council must implement. It can also preserve a narrow emergency path for service-continuity actions, followed by prompt independent review.

Such a design would test the IAB’s concern without pretending that all technical governance decisions should be litigated or that the IAB speaks for every stakeholder.

The evidence to watch is therefore procedural and observable. Before the Council starts, are the appeal rules published and funded? Can the reviewer be appointed and removed independently of the Council? Do designation, removal and performance actions fall within its remit? Are decisions and outcomes reported, with justified confidentiality for security-sensitive material? If the answers remain “to be considered in Establishment”, then the proposed structure is choosing institutional learning first and external recourse later.

That may be defensible—but the Board should explain how an affected party can seek correction during the interval and how early rules can still be changed.