Summary

  • Before October 2016, NTIA was a named customer under a finite federal contract. It could inspect performance, alter future requirements, exercise or withhold renewal options, run a competition and require transition assistance. That did not make the United States the author of every IANA policy, but it did identify the party holding the continuation decision.
  • The transition intentionally removed that national-government position. Names oversight moved to a structure built around PTI, a contract between ICANN and its controlled affiliate, Customer Standing Committee monitoring, complaint and remedial routes, periodic and special IANA Naming Function Reviews, and a formal separation path capable of selecting a successor.
  • The replacement mechanism is real but distributed. A serious naming failure must ordinarily travel through remediation, ccNSO and GNSO approval thresholds, a review team, a separation working group, ICANN Board decisions and Empowered Community safeguards. The naming customers are not third-party beneficiaries of the ICANN-PTI contract, while ICANN is simultaneously customer, funder, sole member and controlling parent of PTI.
  • The post-transition bargain therefore exchanged concentrated state leverage for plural legitimacy. It reduced the risk that one government could use renewal or root-zone authorization as political power, yet it also left no independent counterparty able to say, on its own authority and within a short contractual timetable, that failure requires a new operator.

A transition that removed a principal, not the need for one

On 1 October 2016, a federal contract expired and did not return. That simple legal event completed years of political argument about the United States government's unusual role in the coordination of the Internet's unique identifiers. The transition was celebrated as the completion of privatization and as proof that a global multistakeholder community could supervise critical technical functions without a privileged state. Both descriptions captured an important achievement.

Neither fully answered the institutional question left behind: if the operator later failed, who possessed the authority, standing and practical ability to replace it?

The answer depends on what "replace" means. A customer can complain about a delayed request without replacing the operator. A monitoring body can publish poor performance without holding a termination right. A review team can recommend structural change without signing the next contract. A corporation can possess legal rights against its affiliate while being reluctant to use them. A community can hold powers enforceable against a board while still needing several constituent bodies to agree before those powers become available. The post-2016 settlement contains all of these elements.

None is identical to the former federal customer.

This is not an argument that the transition failed or that NTIA should resume stewardship. The old structure carried a legitimacy deficit of its own. The global users of names, numbers and protocol parameters did not elect the United States Department of Commerce, and other governments reasonably objected to a single country's reserved place in a globally relied-upon system. NTIA itself described the role as temporary.

Its 2014 criteria insisted that any successor arrangement support multistakeholder governance, preserve security and openness, meet the needs of customers, and avoid replacing American stewardship with a government-led or intergovernmental solution.

Yet removing an objectionable principal does not abolish the functions performed by a principal. Someone still has to specify the service, observe performance, demand correction, preserve evidence, fund continuity and make a credible choice when continued incumbency becomes more dangerous than transition. The post-transition design distributed those tasks. Distribution increased representation and made unilateral capture harder. It also separated observation from escalation, escalation from recommendation, recommendation from approval, and approval from execution. The missing counterparty is found in those seams.

What NTIA actually held

Before the transition, NTIA's power was often described too broadly. It did not write all policies for domain names, IP address allocation or protocol parameter registries. The relevant technical and policy communities developed rules that the IANA operator was expected to implement. The federal contract was a procurement for defined functions, not a general licence for the United States to govern Internet content or rewrite community policy. Its value as a control device came from a narrower set of rights.

The contract named the United States government as customer and ICANN as contractor. It set terms, deliverables, security requirements, continuity duties, reporting obligations and inspection rights. Later versions contained service measures and detailed provisions for root-zone management work, number-resource allocations to the regional registries, protocol parameter assignments and other IANA responsibilities. NTIA could assess whether the contractor had performed the specified service even where substantive policy came from elsewhere.

Time gave the arrangement its strongest leverage. The contract did not confer a perpetual franchise. Base periods and government-held options created recurring decisions about continuation. NTIA could revise requirements in a later solicitation, decline to exercise an option, seek competing proposals or extend the incumbent for a limited period while a broader transition was completed. In 2012, it cancelled an initial solicitation after determining that none of the submissions met the stated requirements, then issued a revised procurement and awarded the work.

That episode concerned proposal compliance rather than a finding of operator default, but it demonstrated that designation followed an actual purchasing decision.

The old contract also required assistance with succession. Documentation, deliverables, government rights in certain work products, continuity planning and cooperation with a future contractor mattered because termination without transferability would have been reckless. NTIA's practical leverage was never unlimited: replacing a specialised operator could threaten the stability the government was obliged to protect. Even so, the contract placed the choice and the preparatory rights in one identifiable hand. ICANN knew which party would decide whether the next term existed.

NTIA's separate authorization role in root-zone changes made the government position more visible, but it should not be confused with the whole replacement power. In the pre-transition names chain, ICANN processed requests, NTIA verified and authorized eligible changes, and the root-zone maintainer implemented them. The transactional check was concentrated in names and did not extend in the same form to every number or protocol parameter action. Replacement leverage came primarily from the customer relationship and finite term, while root-zone authorization added a particularly sensitive operational checkpoint.

This combination produced clarity at the cost of asymmetry. One national administration possessed contractual standing for services used everywhere. It could act within the procurement even if other governments, registries or users disagreed with its judgment. The same clarity that made enforcement legible also made the arrangement difficult to defend as a durable basis for global legitimacy.

Why the counterparty mattered even when sanctions were rare

A power does not need frequent public use to shape conduct. The available record does not show NTIA repeatedly threatening default or replacement whenever an IANA measure was missed. That absence is not proof that the contract lacked effect. An operator facing inspection, a dated report, a finite term and a customer able to run another competition has reasons to correct problems before they become constitutional crises.

The crucial quality was not aggression but standing. NTIA did not need to assemble a coalition merely to ask for a contract deliverable. It did not have to persuade several differently constituted communities that the same failure justified the same remedy. It could request records, communicate deficiencies and make a renewal decision in its own legal capacity. A court could identify the parties, the instrument, the duty and the governing procurement rules.

Standing also concentrates accountability for inaction. If a government customer tolerated serious nonperformance, observers knew whom to question. A distributed system can be more inclusive while making omission harder to assign. One body may observe the problem, another may decide it is not yet systemic, a third may agree on the diagnosis but reject the proposed remedy, and a board may prefer one more period of remediation. Each decision may be defensible. Together they can preserve an incumbent through failures that no entity affirmatively chose to accept.

The former arrangement therefore supplied what infrastructure governance often lacks: a single actor that both valued continuity and had a legally intelligible exit option. This did not guarantee wise use. A politically motivated or technically mistaken customer could have abused the same concentration. The transition debate was justified in treating that risk as serious. But when the concentrated actor disappeared, its useful functions had to be rebuilt deliberately rather than assumed to arise from community sentiment.

Three operational communities, three different answers

The phrase "IANA stewardship" can conceal important differences among names, numbers and protocol parameters. Each operational community entered 2016 with its own institutions, customers and expectations. The transition did not produce one uniform successor to NTIA. It preserved or created different legal relationships for each function while placing day-to-day performance in PTI for operational coherence.

For number resources, the five regional Internet registries and ICANN signed a Service Level Agreement for the IANA Numbering Services. That agreement became effective with the transition. ICANN then subcontracted performance to PTI. The regional registries therefore retained a direct agreement with ICANN as the designated operator, with service commitments and review arrangements grounded in the numbering community's own institutions. If performance failed, the numbering parties could look to a contract to which they were signatories, even though the personnel performing the service sat in PTI.

For protocol parameters, the relationship remained anchored in the long-standing memorandum between the Internet Engineering Task Force side of the standards community and ICANN, supplemented for the transition and paired with an ICANN-PTI subcontract. The standards community had an established ability to define registry work through its documents and to reconsider the operator relationship if the service ceased to meet its needs. Again, PTI's performance sat beneath a relationship in which an operational community had a recognised institutional counterpart.

Names were harder. Generic and country-code top-level-domain operators are direct customers, but they do not form one contracting person comparable to the five regional registries. The names community built a dedicated oversight settlement: a new affiliate, a naming function contract, a Customer Standing Committee, service expectations, periodic reviews, special reviews and a separation path. Its sophistication reflects the difficulty of giving a plural community the powers of a customer without creating a new sovereign or transferring the function to a treaty body.

This difference is essential to the vanishing-counterparty thesis. There is no single answer across all IANA services. The clearest gap concerns the naming function, where ICANN became the legal customer of its own controlled affiliate and the actual service customers were granted monitoring and escalation roles rather than ordinary party status under that contract.

PTI isolated the function without creating an external owner

The transition proposal created Post-Transition IANA, incorporated as Public Technical Identifiers, to identify and legally separate the performance of the IANA functions from the rest of ICANN. PTI is a California nonprofit public benefit corporation. Its own board, officers, accounts, audits, plans, contracts and records make the function more visible than an undifferentiated department would be. A successor could, at least in principle, be constructed from a clearer set of responsibilities and assets.

But PTI was never designed as an independent rival to ICANN. ICANN must remain its sole statutory member. As member, ICANN elects all PTI directors, including the two nominees selected through ICANN's Nominating Committee. Three of the five directors are employees of ICANN or PTI, one of them the PTI president. ICANN provides funding and many shared services, and PTI's purpose is to perform the IANA functions on ICANN's behalf. The transition proposal itself described PTI as a controlled affiliate.

This was a deliberate continuity choice. Moving a globally relied-upon service into a fully autonomous institution on a fixed political date would have introduced personnel, systems, finance and authority risks. Keeping PTI within the ICANN family allowed the existing IANA team and operating knowledge to continue. It also avoided building another full multistakeholder corporation around a narrow technical mission. The design sought functional separability without immediate economic or corporate independence.

That choice changes the meaning of the naming contract. ICANN is the customer and PTI is the contractor, but ICANN also controls the contractor through membership, director elections, funding and related agreements. If PTI fails, the party expected to enforce the contract is not an outside purchaser dealing at arm's length. It is the parent that helped appoint the board, finances the service and supplies resources on which performance depends.

The ICANN bylaws try to answer that concern directly: they require ICANN to enforce its rights under the naming contract and protect material terms from easy alteration. Yet a duty to enforce does not create a new independent right-holder. It tells ICANN what it must do; it does not give a registry operator the same ability to sue for the contract remedy that the named customer possesses.

The contract is real, but its right-holder is conflicted by design

The IANA Naming Function Contract is not ceremonial. It imposes service, reporting, inspection, security, continuity, complaint, review and transition obligations. It begins with a five-year term and renews automatically for successive five-year periods unless approved community mechanisms support termination or non-renewal. PTI must cooperate with the Customer Standing Committee, address complaints, analyse failures, produce mitigation plans and maintain a plan for moving the function to a successor.

It also contains a revealing boundary: no provision creates rights, benefits or interests in a third party. The registries that rely on performance are customers in the practical sense and may use specified complaint, mediation and review routes. They are not transformed into beneficiaries entitled to enforce every term as if they had signed the agreement. The two formal parties remain ICANN and PTI.

This creates a structural circularity. ICANN can direct correction, inspect systems, approve certain matters and, after the required approvals, end or decline to renew the contract. PTI can point back to ICANN's own funding and service duties because its performance is expressly conditioned on ICANN providing necessary resources under the Services Agreement. If a failure arises from limited public evidence funding, shared support or parent-level decisions, the nominal customer may be implicated in the contractor's inability to perform.

The contract recognises related-party reality in other ways. ICANN must provide or make available personnel, equipment, services, facilities and funding. Intellectual property created under the naming work belongs to ICANN and is licensed back as needed during the term. ICANN indemnifies PTI while the affiliate relationship continues. These provisions can support continuity and preserve assets for a future transfer, but they also show why the legal line between customer and operator does not create economic independence.

An arm's-length customer can threaten to move its purchase elsewhere while keeping its own organization intact. ICANN cannot replace PTI without changing a core part of how ICANN fulfils its mission, finances the function and manages operational knowledge. The decision is therefore constitutional before it is commercial. That raises the cost of using the remedy and explains why replacement requires more than an ordinary notice of breach.

The Customer Standing Committee observes; it does not appoint a successor

The Customer Standing Committee is the closest permanent body to a collective voice for naming customers. Its core voting membership includes representatives of generic and country-code top-level-domain registry operators, with additional liaisons. It receives performance reports, reviews service levels, engages PTI over concerns and can use remedial procedures when a problem appears persistent or systemic.

This is a major improvement in operational visibility. Under the former structure, public performance information depended substantially on contract reporting and government oversight. The CSC gives direct customers a regular place to compare actual performance with agreed expectations and to distinguish isolated complaints from broader trouble. Because it meets and reports repeatedly, it can detect drift before a five-year constitutional review.

But the CSC is not the post-transition NTIA. The ICANN bylaws expressly deny it authority to initiate a change of operator through a special review. If remediation fails, the committee may escalate to the ccNSO and GNSO. Those bodies then apply their own procedures and decide whether further action is warranted. The monitoring body cannot itself terminate, refuse renewal, commission a replacement competition or bind ICANN to a new operator.

This separation has a principled rationale. A small committee focused on service customers should not be able to displace a critical global operator without wider scrutiny. Registry operators' interests are important but not exhaustive; stability, broader users, governments, technical bodies and the public interest also matter. Replacement deserves a higher threshold than complaint resolution.

The price is that evidence and remedy reside in different hands. The body most continuously familiar with poor service must persuade two policy institutions, each with distinct memberships and voting rules, to treat that service history as a constitutional issue. A severe outage may create rapid agreement. A pattern of degrading quality, opacity, weak succession readiness or repeated near-misses may not. Distributed authority is strongest against dramatic abuse and less decisive against cumulative underperformance.

Review converts evidence into recommendation, not immediate control

Periodic IANA Naming Function Reviews evaluate PTI against the naming contract, the needs of direct customers and the expectations of the wider community. Review teams examine performance, openness, budget transparency, CSC evidence and systemic issues. They can recommend amendments and, at the end of a periodic or special review, determine that a separation exercise is necessary.

The breadth is valuable. Replacement should not depend only on speed measures. A review can consider whether reporting is candid, whether oversight bodies function, whether the contract remains fit for purpose and whether observed problems are symptoms of deeper governance weakness. It can invite public evidence and conduct a more searching examination than monthly monitoring permits.

Yet review is episodic. The first review had to be convened by October 2018, and later reviews follow a multi-year cycle. A special review can respond to serious performance trouble between cycles, but only after remedial procedures have failed, ccNSO and GNSO have considered the result, consultation has occurred and both councils approve the step at high thresholds. This protects the operator from impulsive displacement. It also makes activation dependent on sustained institutional coordination.

The 2025 second review illustrates the distinction between oversight and sanction. It found that PTI operated reliably and efficiently and identified no performance deficiency requiring structural correction. Its recommendations concentrated on contract clarity, transparency and review timing. That is positive evidence about the current operator. It does not prove how quickly the structure would act under a genuinely contested failure, because a mechanism tested in calm conditions may face different incentives when replacement threatens budgets, personnel and established relationships.

A review team also does not sign the successor contract. Its finding can recommend creation of a Separation Cross-Community Working Group. The recommendation then needs supermajority approval from the ccNSO Council, a GNSO Supermajority, consideration by the ICANN Board and the applicable Empowered Community safeguards. Review supplies an authoritative diagnosis. It does not recreate a principal who can act alone on that diagnosis.

Separation is possible, but deliberately difficult

The most important answer to the core question lies in Articles 18 and 19 of the ICANN bylaws. They provide a genuine route by which PTI may cease performing the naming function. A separation exercise can include a request for proposals, selection of another operator, termination or non-renewal of the naming contract, or divestiture or reorganization of PTI.

The route begins with an IANA Naming Function Review Team. If it concludes that separation should be considered, both the ccNSO and GNSO councils must approve creation of a Separation Cross-Community Working Group by their prescribed supermajorities. The ICANN Board then considers the recommendation after public comment. A board refusal faces elevated requirements and community review rights. If the group is created, it investigates the identified problem, consults publicly and can run a competition, assess candidates and recommend the entity with which ICANN should contract.

Its final recommendation again requires ccNSO and GNSO approval, public comment and ICANN Board approval, together with the surrounding Empowered Community protections. When a different operator is duly selected and the decision becomes final, the bylaws direct ICANN to enter a contract on substantially the approved terms. ICANN must absorb specified selection and transition costs, and PTI has continuing duties to support an orderly transfer.

This is not a fictional remedy. It identifies a route, preserves decision records, constrains unilateral board resistance and ends with a mandatory contracting duty. The transition plan requires staff, expertise and cooperation to move service securely. Fundamental-bylaw status makes the architecture harder for an incumbent board to dismantle when threatened.

But the mechanism is designed as a constitutional last resort, not a purchaser's option decision. It requires multiple affirmative institutional acts over time. No single customer can invoke it. The CSC cannot invoke it. The review team can recommend but not execute it. The ccNSO cannot proceed without the GNSO, and the GNSO cannot proceed without the ccNSO. The board remains a necessary decision point, though the Empowered Community can challenge certain refusals or approvals. ICANN ultimately signs the successor contract.

That sequence increases the legitimacy of a replacement and reduces the chance of capture by one faction. It also creates veto points and delay. A capable incumbent can contest whether failures are systemic, whether remediation has truly failed, whether separation is proportionate, whether a candidate is safer and whether the transition itself presents unacceptable risk. Some of that contest is healthy. The analytical question is not whether safeguards should disappear, but whether the distributed chain can maintain momentum when each entity bears transition risk and no independent customer bears sole responsibility for choosing.

The Empowered Community supplies legal force at one level removed

The transition's wider accountability reforms created the Empowered Community, a California nonprofit association through which five ICANN community institutions exercise enumerated powers. It can reject certain PTI governance actions, challenge budget and bylaw decisions, remove directors, invoke review routes and enforce protected rights. This gives community decisions a legal character stronger than advisory consensus alone.

Those powers matter to operator replacement. ICANN cannot freely amend protected PTI governance terms, dispose of key assets, reorganize the affiliate or weaken material features of the naming arrangement without exposure to community action. If the ICANN Board resists duly supported review or separation steps, the community has routes to require reconsideration and, in appropriate circumstances, seek enforceable relief. A board cannot simply erase the replacement architecture when it becomes inconvenient.

The Empowered Community nevertheless is not the service customer under the naming contract. It does not conduct monthly performance management, and it does not hold a free-standing right to choose an operator whenever it sees fit. Its five decisional entities must act through their own procedures and applicable thresholds. Their constituencies have different interests: address policy organizations, country-code managers, generic-name groups, individual-user representatives and governments do not necessarily define operational failure or acceptable transition risk in the same way.

This is legal force without a single principal. It is well suited to stopping an extreme board action that several institutions can identify and oppose. It is less like the former arrangement in which one customer could decide, at a known end date, whether an option should be exercised. The community can compel respect for the constitution; it does not become an arm's-length purchaser simply because its powers are enforceable.

Replacement leverage became asymmetric within the corporate family

After 2016, ICANN possesses formal rights that look familiar: inspection, performance demands, mitigation directions, non-renewal after approved recommendations and the ultimate duty to contract with a successor. Yet the incentives around those rights differ from NTIA's.

NTIA could replace ICANN as contractor while the department remained the department. ICANN replacing PTI would be rearranging its own controlled affiliate, employees, funding, records, intellectual property, facilities and public responsibilities. Even if a new operator were legally separate, ICANN would remain responsible for the contract and for financing transition costs under the approved settlement. The remedy reaches into the customer's own body.

The same dependence can run in reverse. PTI needs ICANN's funding and shared support. If underperformance traces to an ICANN budget choice, a staffing constraint or a service supplied by the parent, insisting that PTI alone cure the failure may misidentify control. The contract requires ICANN to provide necessary resources and makes PTI performance conditional on that support. The parent is both enforcer and a possible contributor to failure.

Corporate law imposes fiduciary duties on PTI directors, including the two directors nominated through a route designed to add outside judgment. That prevents the board from being treated as a mere department head. It does not erase member control. ICANN elects all five directors; three are employees of ICANN or PTI; the PTI president is selected by ICANN as member; and ICANN controls the funding environment. Formal board deliberation can improve decisions while leaving ultimate ownership concentrated.

The design therefore isolates accountability evidence more successfully than it isolates replacement incentives. Separate minutes, audits, budgets, plans and performance reports make it easier to see what the operator does. They do not create a purchaser economically detached from the operator's continued existence.

The transition plan makes succession imaginable, not self-executing

Replacement power is credible only if a successor can take over without damaging the identifier system. The naming contract requires PTI to maintain a transition plan, update it and provide staff and expertise to a successor. The current plan covers the practical categories needed for continuity: service obligations, platforms, people, documentation, security, relationships and the movement of responsibilities.

This is a significant safeguard. An incumbent with exclusive knowledge can make any termination clause unusable. Documenting systems and dependencies reduces that hold-up. ICANN ownership of work created under the naming contract can also help ensure that essential rights remain available for continuity rather than departing with the contractor. The prohibition on PTI subcontracting the whole obligation prevents opaque delegation from hollowing out the operator before a transition.

But a plan is not a standby operator. It does not prove that a qualified alternative can be selected quickly, recruit or retain the necessary people, obtain facilities, satisfy security requirements, establish trust with root-zone partners and perform at global scale. The public record does not offer a live competitive market of interchangeable IANA naming providers. The more specialised and reliable PTI becomes, the harder it may be to compare a successor without putting continuity at risk.

Nor does the plan decide who bears the moment of judgment. PTI prepares it with ICANN input; ICANN and the CSC review it at prescribed intervals; a future separation group would use it in a wider decision. The same institutions whose relationships are being changed must assess readiness. That is unavoidable to some degree, but it is not the independent customer's leverage once represented by an expiring federal award.

Successor readiness should therefore be evaluated through evidence rather than the mere existence of a document. Can key responsibilities be identified without relying on one individual? Are rights, records and security materials transferable under controlled conditions? Can services be separated from shared ICANN support? Are dependencies on the root-zone maintainer, registries and standards bodies mapped? Can a candidate demonstrate capability before the incumbent is displaced? Those questions determine whether the constitutional remedy is operationally usable.

Failure has more than one shape

The replacement structure is easiest to imagine after a spectacular outage. If PTI could no longer process valid root-zone requests, lost the integrity of critical records or suffered an extended service collapse, direct customers and community bodies would probably converge quickly. Evidence would be visible, urgency shared and continued incumbency hard to defend.

More difficult cases are gradual. Performance may remain within narrow measures while transparency declines. Staff concentration may create succession risk without causing a missed service level. Shared costs may rise while responsibility becomes harder to attribute. A series of recoverable incidents may reveal brittleness without producing one decisive breach. Customers may disagree over whether a disputed request reflects poor service or a policy boundary beyond the operator's control.

Political failure is harder still. PTI is supposed to implement established policy without choosing substantive outcomes. If pressure from ICANN leadership, a government or a powerful constituency influenced operational treatment, evidence might be contested and partly confidential. A monitoring committee could see anomalies without having the legal authority to compel every record. The issue might implicate ICANN as much as PTI, weakening the premise that ICANN can cure the problem merely by enforcing against its affiliate.

Financial failure could also originate above the operator. PTI is funded through ICANN and depends on approved budgets and shared services. Persistent underfunding might degrade resilience while each annual plan appears balanced. Rejecting a budget is a community power, but rejection does not itself produce a better funded alternative. Replacing PTI without changing the funding source could move the same constraint to a new contractor.

These cases explain why a single replacement lever and a plural review structure each have advantages. The single customer can act decisively but may misread or politicise a complex problem. The plural structure can test evidence across constituencies but may convert ambiguity into delay. The post-transition settlement chose the second risk because unilateral governmental control was judged more dangerous. That choice should be acknowledged, not obscured by saying that the community simply "replaced" NTIA.

What vanished, and what did not

Operational accountability did not vanish in 2016. PTI publishes extensive performance information. The CSC monitors the naming service. Customers have complaint and mediation routes. Review teams examine the contract and performance. The ccNSO and GNSO possess defined escalation roles. The Empowered Community protects key decisions. PTI must maintain continuity and transition capabilities. The bylaws can compel ICANN to contract with a properly selected successor.

What vanished was the combination of four attributes in one outside institution. NTIA was legally distinct from the contractor. It was the named customer. It controlled a recurring continuation decision. It could prepare a competition and demand handover without first obtaining the approval of multiple bodies inside the contractor's wider governance ecosystem.

No post-transition institution has all four. ICANN is the named customer but not independent of PTI. The CSC is close to customers but lacks the replacement right. The ccNSO and GNSO hold crucial approvals but must act together. Review teams can diagnose and recommend but not appoint. The separation group can run a competition and recommend a candidate but exists only after several gates. The Empowered Community can enforce protected powers but is not the routine purchaser. The eventual successor contract is signed by ICANN.

Calling this a missing "enforceable counterparty" does not mean the structure is legally unenforceable. It means there is no arm's-length principal outside the operator's corporate family that can combine evidence, direction, renewal and replacement in its own capacity. Enforceability has moved upward into constitutional duties and sideways across community institutions.

That movement can be desirable. It prevents one actor from threatening replacement for national advantage. It requires reasons, consultation and public support. It protects the incumbent against a faction that mistakes an unpopular policy outcome for operational failure. The loss is speed, simplicity and a clear owner of the exit decision.

Legitimacy and leverage moved in opposite directions

The transition's strongest achievement was negative power: it removed a privileged state's ability to stand between the operator and the root-zone maintainer and ended a contract whose continuation belonged to one government. This reduced the credible fear that changes in American politics could determine the operator's tenure. It also strengthened the argument that the IANA functions serve global operational communities rather than a national sponsor.

Its strongest replacement safeguard was pluralism. A successor cannot be selected casually. Country-code and generic-name institutions must agree at high thresholds. Review evidence must be developed. Public comment must occur. The ICANN Board must act under protected rules. The wider community can challenge certain decisions. This makes capture by a single government, company or constituency substantially harder.

Pluralism, however, is not free. Each additional veto protects against one kind of abuse while enabling another kind of stasis. High thresholds make a replacement more legitimate after agreement is achieved, but they lower the probability that disagreement will produce any replacement at all. The incumbent benefits from uncertainty because continuity is itself a powerful value. When the cost of a false positive is a dangerous transition and the cost of a false negative is gradual decline, institutions tend to wait.

The bargain can therefore be stated precisely. The old system offered high clarity of leverage and low global representativeness. The new system offers broader representativeness and lower concentration of leverage. It did not eliminate power; it decomposed power into observation, escalation, review, approval, enforcement and contracting. Whether that is enough depends on how well the pieces join under stress.

A stronger post-transition discipline does not require restoring government control

The missing counterparty cannot be solved simply by appointing another state or intergovernmental body. That would reverse the transition's central legitimacy gain and recreate geopolitical control in a different form. Nor should a small customer committee receive an unrestricted power to dismiss the operator. The continuity and capture risks are too great.

A better discipline would make the existing distributed authority more usable. First, performance evidence should include successor readiness, dependence on shared ICANN services, staffing concentration, recoverability and control over essential rights - not only completion-time measures. A service can meet monthly targets while becoming harder to transfer or more dependent on one corporate parent.

Second, the boundary between PTI-caused and ICANN-caused failure should be explicit. If funding, facilities or shared support contribute to a problem, remediation should address the party controlling that resource. Otherwise the customer-parent can appear to enforce a contract while leaving its own decisions untouched. Review teams need enough evidence to attribute responsibility across the relationship.

Third, escalation should preserve a public chronology. The date a problem was identified, attempted remedies, reasons for closure or continuation, council decisions and unresolved dependencies should be legible without exposing security-sensitive material. That record makes delay accountable. It also helps distinguish prudent caution from institutional avoidance.

Fourth, transition exercises should test bounded capabilities without pretending that a full replacement can be rehearsed. Controlled demonstrations of records transfer, contact handover, access revocation, service reconstruction and candidate due diligence can expose assumptions in the transition plan. The goal is not to manufacture a crisis but to keep the remedy credible enough that an incumbent cannot rely on irreplaceability.

Finally, each institution should state its role plainly. The CSC monitors and remedies; it does not select. Review teams diagnose and recommend; they do not contract. The ccNSO and GNSO authorize crucial steps; they do not operate the service. The separation group can assess candidates. The Board makes decisions subject to protected community powers. ICANN signs and funds the eventual contract. Clarity about division of labour is the closest a plural settlement can come to the accountability once supplied by one named principal.

Conclusion: a safer system with a longer hand on the lever

Who can replace the IANA naming operator after 2016? No single institution can do so at will. A successor can emerge only through a protected chain: operational evidence, failed remediation where a special review is needed, approval by country-code and generic-name bodies, a review recommendation, creation and work of a separation group, Board action, Empowered Community safeguards and an ICANN contract with the selected provider. PTI's transition duties are meant to make the result safe.

That chain is an answer, but it is not the answer NTIA once represented. The federal customer combined distance from the operator, clear legal standing, a finite term and practical authority over continuation. The transition removed that concentration because it was nationally asymmetric and vulnerable to political misuse. In its place came a more globally defensible constitution whose replacement authority is shared and deliberately slow.

The post-transition system should be judged by both sides of that bargain. It reduced state control and insulated the root from one government's changing preferences. It also made replacement dependent on coalition, procedure and action by ICANN against an operator ICANN owns and funds. Present performance has been strong enough that the hardest parts have not been tested by a sustained failure. That is welcome, but it leaves a question rather than proof.

The vanished counterparty is therefore not a nostalgic symbol of American stewardship. It is a governance function: the independent person that can receive evidence, require performance and choose another provider. After 2016, that function survives only when several institutions act as one. The central resilience test is whether they can do so before continuity becomes an argument for preserving the very operator whose failure made replacement necessary.

Sources