Summary

  • ICANN's Work Stream 2 jurisdiction recommendations are best read as a legal-output file, not as a referendum on the geography of Internet governance. The official implementation page identifies two operative families: recommendations about sanctions and recommendations about choice of law and venue in ICANN agreements.
  • The strongest concrete change concerned affected applicants and contracted parties. ICANN moved toward best-efforts licensing obligations, clearer communication, and guidance that a non-United States registrar does not become bound by OFAC sanctions merely because it has a registrar agreement with ICANN.
  • The project did not move ICANN away from California. ICANN's amended and restated articles, filed in California on 3 October 2016, still identify it as a California nonprofit public benefit corporation, and current bylaws still carry the California nonprofit public-benefit heading.
  • Participation should therefore be evaluated by enforceable rights: who gained a contractual claim, who gained a venue option, who gained a clearer sanctions path, and who still had only the ability to comment, entity, or persuade. Counting meetings, filings, comments or pages cannot answer that question.

The legal question was smaller than the political argument

The phrase "ICANN jurisdiction" can sound like a constitutional question about where Internet governance belongs. In the public imagination it easily becomes a fight over whether a global identifier coordinator should be subject to United States law, California courts, United States sanctions, Los Angeles arbitration clauses, and the habits of an American nonprofit corporation. That wide framing is understandable. ICANN coordinates parts of a global technical system. Its contracted parties, users, governments and critics are spread across many legal orders.

A corporation with global reach but a California legal home will always generate unease when the laws of that home create practical consequences for people elsewhere.

The Work Stream 2 jurisdiction project was narrower. It did not create a world corporation. It did not transfer ICANN to treaty status. It did not replace California nonprofit law with a global public-law code. It did not give every affected Internet user a court in their own country. The official jurisdiction implementation page describes recommendations about two operational categories: trade restrictions, especially United States OFAC sanctions, and choice of law or venue provisions in ICANN agreements. Those are important subjects.

They affect who can become a registrar, who can apply for a new generic top-level domain, how a contracted party understands governing law, and where some disputes may be heard. But they are not the same as moving the corporation.

That distinction is the core of the article. The jurisdiction project matters precisely because it shows the boundary between participation and enforceable legal alteration. The community could identify frictions. The subgroup could write recommendations. ICANN org could implement some of them. Contracted parties could receive clearer clauses or options. Applicants in sanctioned jurisdictions could receive a stronger best-efforts promise.

Yet the main legal architecture remained recognizably the same: ICANN was still the California corporation, acting through bylaws, articles, contracts, board resolutions, internal accountability mechanisms and external law.

This is not a criticism of every jurisdiction recommendation. Some were useful because they made a vague risk more concrete. A sanctioned-country applicant does not need an abstract debate about sovereignty as much as a clear answer on whether ICANN must seek a license and communicate during the attempt. A non-United States registrar does not need a philosophical claim about international law as much as a warning that a copied contract clause may overstate its obligations. A registry negotiating an agreement does not need a global parliament as much as a known menu of legal choices and a dispute venue that can be evaluated before signing.

These are real gains.

The mistake is to confuse those gains with constitutional relocation. The corporation did not move because the recommendations did not create the legal machinery required to move it. They operated at the level of obligations, choices, clarifications and implementation commitments. They made parts of the California-centered contract system less blunt. They did not replace that system.

The corporation remained where the articles put it

The strongest evidence is formal. ICANN's amended and restated articles of incorporation state that the president and secretary certified the corporation as a California nonprofit public benefit corporation. The document was approved by the ICANN Board on 9 August 2016 and filed with the California Secretary of State on 3 October 2016, just after the IANA stewardship transition took effect. It states that the corporation is not organized for private gain and is organized under the Nonprofit Public Benefit Corporation Law for charitable and public purposes.

It also ties the corporation's mission to the bylaws, the Internet community as a whole, and open and transparent activity.

That filing is not a footnote. It is the legal container in which the rest of the accountability reforms sit. The post-transition settlement created new community powers, mission limits, review mechanisms and separation structures, but it did so inside a California nonprofit public benefit corporation. The articles also say the corporation has no members, while separately requiring Empowered Community approval for specified changes. That arrangement is a legal design choice: community power is attached through a separate mechanism rather than by converting ICANN into a conventional membership corporation with members in every jurisdiction.

The bylaws preserve the same corporate premise. The current official bylaws page is headed as the bylaws for the Internet Corporation for Assigned Names and Numbers, a California nonprofit public-benefit corporation. The bylaws define mission, powers, transparency, accountability review, reconsideration, independent review, the Empowered Community, board composition, supporting organizations and advisory committees. They also state that ICANN shall not act outside its mission and that it does not hold general governmentally authorized regulatory authority.

Those mission limits are meaningful. They narrow what ICANN may do and create review hooks when actions exceed the written bargain. But a mission limit is not a relocation clause. It constrains the California corporation; it does not dissolve it. It may protect a requester from a mission breach, but it does not turn the requester into a shareholder, a treaty beneficiary or a plaintiff with automatic standing in a preferred national court. The legal question therefore has to be asked at the right level. Did the jurisdiction project change the corporation's domicile? No.

Did it alter some obligations in the contracts and applications through which the corporation interacts with others? Yes.

That answer can disappoint both sides. Critics who wanted ICANN to escape United States legal exposure will find the result thin. Defenders who prefer to treat jurisdiction criticism as empty politics should also be careful. The project produced concrete admissions that United States law and California-centered contract design can impose real frictions on global parties. The fact that the corporation did not move does not mean the problem was imaginary. It means the chosen remedy was targeted rather than structural.

OFAC is where the project became most concrete

The sanctions recommendations are the clearest example of legal outcome rather than governance theatre. The jurisdiction page identifies Recommendation 4.1 and its components as relating to the impact of trade regulations, including those from the United States Office of Foreign Assets Control, on whom ICANN may provide with goods and services. That framing is practical. Sanctions are not just political atmosphere. They can prevent ICANN from entering a registrar accreditation agreement, processing a new gTLD application, supporting an applicant, funding travel or providing another service when United States law restricts the transaction.

The first component addresses registrar accreditation applicants from sanctioned countries. The page explains that ICANN would need an OFAC license to enter a registrar accreditation agreement with such an applicant. It records a prior uncertainty: ICANN had not been obligated to seek such licenses, and OFAC might decline a requested license. The recommendation was that ICANN's terms should be amended so ICANN would apply for and use best efforts to secure a license if the other party was otherwise qualified and not individually sanctioned. It also called for helpful and transparent communication during the licensing effort.

The implementation status shown on the page marks that component completed in Q1 2022.

This is a real right, but it is a limited one. It does not guarantee accreditation. It does not force OFAC to issue a license. It does not make United States sanctions disappear. It does not give an applicant a general exemption from law. It changes the conduct expected of ICANN: apply, use best efforts, communicate, and do so when the applicant is otherwise qualified and not individually subject to sanctions. The right is procedural in form but substantive in effect because it changes the default from optional pursuit to a required effort.

The second component addresses new gTLD applicants from sanctioned countries. The 2012 round created practical barriers for residents of sanctioned countries to move through the application sequence. The recommendation again focused on ICANN's commitment to apply for and use best efforts to secure OFAC licensing where an applicant would otherwise be approved and is not on the specially designated nationals list. The official page marks this completed in Q1 2023. Again, the legal outcome is concrete but bounded. It gives the applicant an ICANN-side obligation to try.

It does not eliminate a sovereign licensing decision by the United States government.

This is exactly the right lens for measuring the jurisdiction project. If a person asks whether the project internationalized ICANN, the answer is no. If a person asks whether it created a more enforceable path for a qualified registrar or gTLD applicant affected by sanctions, the answer is closer to yes. The difference is not semantic. It determines who can rely on the result. A entity in a public debate cannot sue merely because the debate occurred. A qualified applicant may be able to point to a changed term, implementation document, or commitment and ask whether ICANN used best efforts in the promised way.

The non-United States registrar problem shows why clarity matters

The third OFAC component is less dramatic but institutionally important. The official page says some non-United States registrars appeared to apply OFAC sanctions to registrants or potential registrants based on a mistaken assumption that having a contract with ICANN required them to comply with OFAC sanctions. It also notes that some may have copied registrant agreements from United States-based registrars. The recommendation was not that ICANN become the legal adviser for every registrar.

The recommendation was that ICANN clarify that the mere existence of a registrar accreditation agreement does not itself require non-United States registrars to comply with OFAC sanctions, and that ICANN explore tools to remind registrars to understand applicable law and accurately reflect it in customer relationships. The official page marks this completed in Q4 2023.

This recommendation is a small but valuable illustration of jurisdictional overhang. Law can travel not only through command but through imitation. A non-United States registrar may over-comply because it fears ICANN, misunderstands a contract, copies another registrar's customer terms, or sees United States legal language as a safe default. The harm then appears outside the formal reach of the law. A customer is refused, a registrant is chilled, or a market treats a United States compliance standard as global because the institutional center is in California and the contract counterparty is ICANN.

The recommendation addressed that diffusion problem by clarification. It did not declare that non-United States registrars never face sanctions duties. Their own governing law, payment providers, ownership, customers, banks and operations may create obligations. It did not invite registrars to ignore applicable law. It instead separated an ICANN contract from a universal OFAC command. That separation matters because many accountability failures begin when institutions allow cautionary myths to harden into private rules. A contract with ICANN can become a reason for customer exclusion even when the law does not require that exclusion.

The legal outcome remains limited. Clarification does not compensate affected registrants. It does not create a fast complaint route for every erroneous denial. It does not require publication of registrar-level sanctions decisions. It does not standardize every customer agreement. But it changes the information environment in which registrars draft and enforce terms. For a global coordinator, that is often the realistic level at which jurisdiction reform can work: not replacing national law, but preventing the coordinator's own documents from exaggerating the reach of one national law.

General licenses show the hardest boundary

The fourth sanctions recommendation is the most revealing because it remained, on the official page reviewed for this article, in progress with a targeted Q2 2026 completion date. It concerns OFAC general licenses. A general license can authorize a class of transactions without requiring a specific license for each case. The recommendation suggested that ICANN take steps to pursue one or more general licenses for transactions integral to its role in managing the DNS and contracts for Internet resources. It also recognized that developing such a license would require work with the United States Department of the Treasury and regulatory action.

This is where the jurisdiction project runs into the hardest boundary. ICANN can apply. ICANN can study. ICANN can report obstacles. ICANN can seek community advice. It can communicate progress and identify other ways to reduce friction if unsuccessful. But ICANN cannot unilaterally write the United States sanctions regulations. A global community recommendation can become an ICANN obligation to try; it cannot become an order to the Treasury Department.

The distinction is essential for judging participation. A person who joined the jurisdiction debate may have believed that the problem was United States legal exposure. That belief can be accurate. But the remedy path depends on who controls the relevant legal lever. Where the lever is an ICANN application term, ICANN can change it. Where the lever is an ICANN contract, ICANN can amend or offer choices. Where the lever is a United States regulatory license, ICANN can request and advocate but cannot decide. The public participation system should be honest about this boundary at the beginning, not only when a recommendation stalls.

The general-license recommendation also shows why "friction" is an limited public evidence word unless converted into rights. Friction for whom? A registrar applicant who cannot obtain accreditation? A registry applicant whose application is delayed? A traveler who cannot receive funded support? A privacy or proxy service? A registrant denied service by a cautious intermediary? Each case has a different legal relationship and remedy.

A general license might lower friction across classes, but without a granted license the affected party still needs a fallback: written reasons, evidence requirements, expected time frames, alternate transaction paths, and a way to contest delay or refusal.

The most important outcome may therefore be diagnostic. The jurisdiction project exposed which problems ICANN can fix by its own documents and which require another sovereign actor. That exposure should make future participation more precise. Instead of asking whether "the community" wants less United States legal influence, a serious reform proposal should ask which clause, license, filing, venue, data location, court route or statutory obligation must change, and who has legal power to make that change.

Choice of law is not the same as changing the corporate home

The second recommendation family, 4.2, concerns choice of law and choice of venue provisions in ICANN agreements. This is the part most likely to be mistaken for relocation because it uses the language of governing law. The official page identifies several possible approaches for registry agreements, with similar options for registrar accreditation agreements: a menu approach, a fixed California or United States law approach, a carve-out approach, a bespoke approach based on the registry operator's governing law, and a status quo approach.

The menu approach is particularly important. Under that model, the registry would choose the governing law before execution from a defined menu of possible laws. The page says the subgroup did not determine the specific menu, but discussed options such as one country or a small number of countries from each ICANN geographic region, plus the status quo, the registry's jurisdiction of incorporation, or countries in which ICANN has physical locations. The recommendation was that the registry choose from the options on the menu rather than negotiate the choice with ICANN.

That design can shift leverage at the contract level. It gives a registry operator a structured option. It can reduce the sense that every contractual relationship must be centered on California law or on a law chosen unilaterally by ICANN. It can also improve predictability because the choice is known before execution and bounded by a menu rather than improvised in a dispute.

But a contract-law menu is not a corporate migration. ICANN remains ICANN. Its articles remain California articles. Its bylaws remain the bylaws of a California nonprofit public-benefit corporation. Its board duties and corporate powers remain anchored in that legal form. A registry agreement can use a different governing law for particular contractual obligations while the corporation itself remains subject to California corporate law and applicable United States law. A choice-of-law clause is not a teleportation device.

This matters because public debates sometimes treat jurisdiction as a single container. In reality, jurisdiction is layered. There is the law governing ICANN's corporate existence. There is the law governing a contract. There is the venue or seat for arbitration. There is the law that applies mandatorily because of sanctions, tax, data protection, competition, consumer protection or court order. There is the practical law of banks, payment providers and insurers. There is the law that a court will apply when asked to enforce an award or order. Changing one layer may leave the others intact.

The Work Stream 2 recommendation recognized that reality more than its political slogan did. The menu, fixed-law, carve-out, bespoke and status-quo options are contract tools. The carve-out approach, for example, would keep parts of a contract that need uniform treatment under a predetermined law, possibly California, while letting other parts follow the registry's jurisdiction or a menu choice. That is a sophisticated design for layered legal risk. It is also evidence that the project did not settle on simple relocation as the answer.

Venue optionality is a narrower right than global justice

Choice of venue is also real but narrow. The official jurisdiction page says that registry agreement disputes were resolved through binding arbitration under ICC rules and that the registry agreement contained a venue provision naming Los Angeles, California as both the physical place and the seat of arbitration. The recommendation contemplated offering a list of possible arbitration venues so a registry entering an agreement could choose its preferred venue at or before execution. The page marks this component completed in Q1 2023.

For a registry operator, venue can matter. It affects travel, counsel selection, procedural familiarity, perceived neutrality, local court supervision over arbitration, enforcement strategy and cost. A venue outside Los Angeles may reduce the symbolism and expense of having every serious dispute return to ICANN's home jurisdiction. It may also make it easier for some operators to accept the agreement as a global commercial instrument rather than an American institutional instrument.

Yet venue optionality does not give every affected party a remedy. It belongs to the parties to the agreement. Registrants, civil-society organizations, governments, users and competitors are usually not parties to the registry agreement. They may be affected by the registry's conduct, ICANN's decision, or a public-interest commitment, but they do not automatically gain the right to choose arbitration venue. A remedy that changes the place of contractual arbitration is therefore a contracted-party remedy, not a universal public accountability remedy.

This is why legal-outcome analysis is harder and more useful than participation analysis. If the question is "who participated," the jurisdiction project appears broad. Governments, contracted parties, community groups, lawyers, staff, observers and volunteers could all take positions. If the question is "who gained a legally usable change," the map is narrower. A registry may gain a venue option. A registrar applicant from a sanctioned country may gain an ICANN best-efforts obligation. A non-United States registrar may receive a clarification.

An ordinary registrant may gain only indirect benefit if the registrar or registry acts differently.

That narrower answer is not cynical. It protects the value of participation by refusing to exaggerate it. A participation system is more credible when it says, "This consultation can change these instruments, for these parties, in these ways," than when it invites broad grievances and then delivers clauses usable only by a subset of legal counterparties. People can still participate on principle, but they should know the legal channels through which their input might become enforceable.

Process volume is the wrong metric

The jurisdiction debate is a warning against measuring accountability by institutional activity. A large record can show that a topic was taken seriously. It can also hide the absence of legal displacement. Meetings, comments, subgroup reports, minority statements, implementation pages and status labels are not irrelevant; they are evidence that the institution heard and processed the issue. But they are not the outcome. The outcome is the change in rights, duties, standing, venue, law, deadlines, records, remedies or review.

Participation volume can even distort the evaluation. A heavily attended debate about moving ICANN may produce fewer enforceable rights than a smaller drafting change to an application term. The first may be politically dramatic but legally aspirational. The second may give a real applicant a claim when ICANN fails to seek a license. If the audit counts hours and submissions, the dramatic debate looks larger. If the audit counts rights, the drafting change matters more.

This distinction is especially important in ICANN because many accountability channels are open but indirect. A person can comment, attend, join a constituency, write to a board member, join a working group, file a reconsideration request if materially affected and within scope, support an Empowered Community petition through a decisional entity, or pursue an independent review if standing and subject matter allow it. These are not identical powers. They should not be collapsed into a single word like participation.

The right metric for the jurisdiction project has at least five columns. First, identify the affected class: registrar applicant, new gTLD applicant, non-United States registrar, registry operator, registrant, community entity, government or ICANN itself. Second, identify the legal instrument: terms and conditions, application guidebook, registrar agreement, registry agreement, bylaws, articles, implementation documentation or external law. Third, identify the changed duty: best efforts, communication, clarification, menu choice, venue choice, reporting, study or no change.

Fourth, identify the remedy if the duty is breached: contract claim, accountability mechanism, public escalation, arbitration, court enforcement, or no direct remedy. Fifth, identify the remaining dependency: OFAC licensing, board discretion, contract negotiation, external law, community threshold or cost.

When that table is filled in, the project looks useful but limited. It is strongest where the affected party is a direct ICANN counterparty or applicant and the changed duty can be stated in contract-like language. It is weakest where the affected party is a broad public, a non-party user, or a person harmed by a private over-compliance decision downstream. That is not a reason to dismiss the project. It is a reason to stop describing it as a general cure for jurisdictional legitimacy.

The California anchor was not an accident after the transition

The timing makes the result more revealing. The amended and restated articles were approved and filed in 2016, the same year the IANA stewardship transition ended the historic United States government contract. If there had been a moment when ICANN might have been re-founded in a radically different legal form, that was the moment when the community was redesigning accountability to satisfy transition conditions. The result was not legal homelessness. It was a California corporation with stronger community powers, clearer mission language and new or revised accountability mechanisms.

That choice had reasons. California nonprofit law provides a known corporate form, fiduciary duties, filing records, judicial background, charitable-purpose language and a legal system able to host enforceable rights. Existing contracts, staff, offices, tax status, counsel, insurance, assets and institutional memory were tied to the United States. Moving ICANN would not have been a symbolic address change. It would have required new incorporation, asset transfer, tax analysis, contract assignment, employment changes, regulatory review, governance redesign, litigation risk and acceptance by counterparties.

The legal cost of relocation was always much higher than the rhetorical cost of demanding it.

The jurisdiction project therefore faced a practical choice. It could try to reopen the entire corporate settlement, or it could identify the specific ways California and United States law created operational problems and reduce those problems. The final recommendation family chose the second route. It did not ask whether every entity preferred California. It asked which sanctions and contract clauses caused practical harm and what ICANN could change without destroying the post-transition structure.

That choice is defensible, but it should be named. A targeted repair is not the same as a completed legitimacy settlement. If a party believes no global technical coordinator should be a California nonprofit, the Work Stream 2 outcome did not satisfy that objection. If a party believes the main problem is unnecessary legal friction for otherwise qualified applicants and contracted parties, the outcome was more responsive. A serious evaluation must say which objection is being measured.

The articles also complicate the sovereignty critique. They acknowledge the Internet as an international network of networks and instruct the corporation to pursue global public interest through an inclusive bottom-up multistakeholder community process. They call for conformity with relevant principles of international law and applicable local law. That language is globally oriented, but it sits inside a California filing. ICANN's design is therefore not purely domestic or purely international. It is a domestic corporation charged with a global public-interest mission and constrained by global-facing commitments.

The jurisdiction project adjusted the edges of that hybrid. It did not turn the hybrid into a treaty body.

Affected parties do not all stand in the same legal position

The jurisdiction debate becomes clearer once affected parties are separated. A registrar accreditation applicant from a sanctioned country stands in one position. It is seeking a direct contractual relationship with ICANN. If ICANN's application terms require best efforts to obtain an OFAC license, the applicant has a concrete expectation about ICANN's conduct. The applicant can ask whether it was otherwise qualified, whether it was individually sanctioned, whether ICANN applied, whether ICANN communicated, and whether the effort was genuine.

A new gTLD applicant from a sanctioned country stands in a similar but not identical position. The application may involve different fees, evaluation steps, objections, delegation requirements and business commitments. A license effort can remove one barrier while leaving others. The applicant benefits from a best-efforts commitment, but not from guaranteed approval.

A non-United States registrar stands in another position. It may need clarity that its ICANN contract does not by itself require OFAC compliance. But its own law, banking relationships, corporate ownership, customer base and risk policy may still lead it to screen customers. ICANN's clarification helps prevent contract-based overstatement; it does not rewrite the registrar's entire legal environment.

A registry operator negotiating a registry agreement stands in a contract-design position. A menu of governing laws or arbitration venues may influence bargaining and dispute expectations. But it does not free the operator from mandatory law, nor does it guarantee that every dispute will be convenient or inexpensive. It also does not give non-parties direct control over the venue.

A registrant, civil-society group or ordinary Internet user stands in the weakest direct position. They may benefit indirectly when registrars avoid over-compliance, when applicants from sanctioned jurisdictions can participate, or when registry contracts feel less California-centered. But unless the person has a direct claim under a contract, bylaw mechanism or external law, the jurisdiction recommendations do not give them an individual order power.

This differentiation is the heart of participation boundary analysis. A global consultation can include many affected voices, but the legal output can privilege those with direct contractual privity, application status or institutional standing. That is not necessarily improper. Law often moves through defined relationships. But governance communications should not imply that every entity receives the same enforceable benefit.

The menu approach carries its own risks

The choice-of-law menu looks attractive because it decentralizes a decision that otherwise appears to belong to ICANN. But a menu is only as fair as its contents, timing and consequences. If the menu is too narrow, it can become cosmetic. If it is too broad, it can fragment interpretation of the base agreement. If choices are made by sophisticated registries but not realistically understood by smaller or newer operators, the menu can reproduce expertise inequality. If ICANN controls the menu without transparent criteria, the reform simply moves discretion from one clause to another.

The official page acknowledges the need for balance. Different governing laws applied to the same base agreement can create advantages and disadvantages. Uniform treatment supports consistency, predictability and easier administration. Local or regional law can improve acceptance, neutrality and alignment with a registry's own legal environment. The carve-out approach responds to this tension by preserving uniform law for contract parts that need common treatment while letting other parts vary.

The governance standard should be practical. ICANN should publish the reasons a jurisdiction is on or off the menu. It should explain whether menu choices are based on geographic balance, legal maturity, arbitration support, neutrality, language, court enforceability, nonprofit compatibility, data protection, sanctions exposure or the presence of ICANN offices. It should state whether a registry can change its choice on renewal, whether existing registries can opt in, whether the menu affects public-interest commitments, and whether dispute data will be reported by chosen law and venue.

Without those details, a menu can satisfy the vocabulary of choice without proving the reality of choice. A registry may choose a venue because all available options are expensive. A smaller operator may accept the default because it lacks legal resources to evaluate alternatives. A public-interest claimant may still be outside the contract. The article's point is not that the menu is bad. It is that the menu must be judged as a legal instrument, not as a participation symbol.

This also preserves the value of uniformity where uniformity is justified. ICANN's contracts are part of a global coordination system. Some clauses need common meaning if security, stability, transition, data access or compliance obligations are to be administered consistently. The problem with California law was not that any uniform law is illegitimate. The problem was that one uniform law, tied to ICANN's home, had become a standing symbol of unequal legal distance. A good menu should reduce unequal distance without making the base agreement incoherent.

What relocation would have required

The jurisdiction project did not move the corporation because relocation would have required a different agenda. First, it would have required a destination legal form. Would ICANN become a nonprofit corporation in another country, an international organization, a foundation, a treaty-created entity, a distributed corporate group, or a new parent with a California subsidiary? Each form has different duties, tax treatment, immunities, reporting obligations, board liabilities and court exposure.

Second, it would have required asset and contract transition. ICANN holds agreements with registries, registrars, vendors, staff, office lessors, insurers, technical service providers and related entities. Moving the corporate home could trigger consent rights, tax consequences, assignment restrictions, employment changes and litigation. A relocation plan would need a full continuity design for every critical agreement.

Third, it would have required accountability redesign. The Empowered Community, bylaws, articles, board duties, independent review, reconsideration and community powers are built around the current corporation. A different legal form might strengthen some powers and weaken others. A treaty body could gain international symbolism while reducing private enforceability. A foundation in another jurisdiction could reduce United States symbolism while creating new local law dependencies. A distributed parent could lower single-jurisdiction concentration while making responsibility harder to trace.

Fourth, it would have required a sanctions and mandatory-law analysis. Leaving California would not eliminate every United States legal exposure if ICANN maintained United States operations, United States counterparties, dollar transactions, United States persons, United States-origin services or activities with United States nexus. Nor would it eliminate other sanctions regimes. The result might be multi-jurisdictional compliance rather than freedom from jurisdiction.

Fifth, it would have required community consent on acceptable risk. Some entities might prefer a less United States-centered ICANN even at the cost of uncertainty. Others might prefer the known enforceability of California law over a new home with weaker transparency, unstable courts, stronger governmental control or uncertain nonprofit duties. A real relocation proposal would need to compare these risks, not merely rank jurisdictions by political symbolism.

Work Stream 2 did not do all of that, and the official recommendations should not be judged as though they did. They chose targeted repair. That choice can be evaluated on its own terms. Did targeted repair reduce sanctions licensing uncertainty? In part. Did it clarify over-compliance by non-United States registrars? Yes, at least at the level of ICANN communication. Did it provide contract choice or venue optionality? The official page marks the relevant recommendation completed. Did it move the corporation? No. Did it create a universal public court for ICANN decisions? No.

The legal-outcome audit should become standard

The jurisdiction project suggests a standard for future ICANN accountability debates. Every major consultation should publish a legal-outcome audit after implementation. The audit should not be a celebration of engagement. It should identify the exact documents changed, the parties who can rely on the changes, the deadlines, the enforcement route, the remaining discretion and the known non-parties.

For the jurisdiction project, such an audit would say that certain sanctions-related recommendations affected application terms and ICANN's conduct toward otherwise qualified applicants. It would say that non-United States registrar clarification affected understanding of contractual effect but not the registrars' independent local law duties. It would say that choice-of-law and venue recommendations affected registry or registrar agreements, subject to the final implementation details. It would say that the corporation remained a California nonprofit public benefit corporation under its articles and bylaws.

The audit should also name non-results. No corporate relocation occurred. No general exemption from United States sanctions was created by ICANN alone. No universal right of registrants to choose forum was created. No guarantee of OFAC licensing was created. No automatic remedy for every private over-compliance decision was created. Naming non-results is not hostile. It prevents the institution from converting effort into a claim of cure.

Outcome audits also help entities decide where to spend attention. If a proposed recommendation can only produce a study, affected parties may still support it, but they should not confuse it with a right. If a proposed clause gives a direct applicant a claim, it may deserve more drafting scrutiny than a broad statement of principle. If a recommendation requires action by an external government, the community should know which actor must decide and what ICANN can do if the actor declines.

This discipline would improve ICANN's legitimacy more than another general promise to listen. The problem is not that ICANN lacks channels for talk. The problem is that the conversion of talk into legal effect is often hard to see. The jurisdiction project is one of the clearest files because the difference is stark: thousands of words about global legitimacy ended in a limited set of contract and licensing changes while the corporation stayed put. That contrast should become a teaching example, not an embarrassment.

Participation still mattered, but only inside its legal container

It would be wrong to conclude that participation failed because relocation did not occur. Participation can clarify harms, force institutions to publish reasons, change terms, create commitments and expose false assumptions. The sanctions recommendations likely would not have been as concrete without affected parties and advocates explaining how uncertainty chilled applicants from sanctioned jurisdictions. The non-United States registrar clarification likely would not have had the same force without evidence or concern that ICANN contracts were being over-read downstream.

Choice-of-law and venue recommendations required legal imagination from parties who understood the costs of a single default.

But participation is not magic. It has to enter a legal container. A recommendation becomes enforceable only when it is incorporated into a term, bylaw, policy, procedure, contract, board resolution, or external legal act that defines who can invoke it and how. Otherwise it remains a record of institutional aspiration. Aspiration can shape future decisions, but it does not itself decide a dispute.

The jurisdiction project therefore sets a healthy boundary for the ICANN community. A entity can ask for relocation. A subgroup can study legal risks. A report can recommend narrower repairs. ICANN can implement those repairs. The final evaluation should not ask whether everyone was allowed to speak. It should ask what speech changed. If the answer is "best efforts for licenses, clearer sanctions guidance, and law or venue options," then the project should be praised for those changes and not credited for more.

That boundary also protects ICANN from impossible expectations. A California nonprofit cannot promise immunity from United States law. It cannot guarantee that a sovereign regulator will issue licenses. It cannot make every registrar's local compliance analysis vanish. It cannot make all entities legal counterparties. If ICANN communicates those limits clearly, it reduces disappointment and focuses reform energy on documents it can actually change.

The boundary also protects affected parties from symbolic satisfaction. A community member harmed by sanctions over-compliance does not need to be told that the community debated jurisdiction at length. They need to know whether a registrar's refusal was required by law, required by contract, chosen by policy, copied from another template, or based on misunderstanding. That is an outcome question. The jurisdiction project provides some tools for answering it, but only if those tools are used as legal instruments rather than as proof that the debate is over.

Why this matters for number-resource governance

The article sits in an ICANN category, but the boundary matters to number-resource governance because ICANN also coordinates the top-most level of Internet Protocol numbers and Autonomous System numbers in service of its mission, as requested by the IETF and the RIRs and through related tasks agreed with the RIRs. The bylaws make that role narrower than ICANN's role in generic names. They also exclude disputes relating to Internet numbering resources from reconsideration. Those details prevent the jurisdiction debate from being lazily transferred into every number-resource dispute.

If a number-resource holder has a dispute with a regional Internet registry, ICANN's California corporate seat may be relevant only indirectly or not at all, depending on the function, document and parties. The registry's own jurisdiction, membership terms, policy process, transfer rules, court orders, sanctions duties and service commitments may matter more. Conversely, when a dispute concerns the IANA numbering-services relationship, global policy development, top-level allocation or an ICANN bylaw duty, ICANN's legal form may become more relevant.

The jurisdiction project teaches a method: do not start with the map; start with the legal instrument. Which entity made the decision? Under which document? Who is bound? What is the remedy? Which court or arbitral body can enforce it? Which law is mandatory regardless of contract? Which community process can change the rule? Only after those questions are answered should the word jurisdiction be used.

That method is especially important as number-resource governance becomes more economically consequential. IPv4 scarcity, transfers, leasing, RPKI, reverse DNS, sanctions screening, bankruptcy and collateral all involve legal relationships. A broad claim that "ICANN is in California" does not solve those relationships. A broad claim that "the community participated" does not solve them either. The enforceable outcome depends on the particular contract, registry policy, recognition mechanism, court order or operational credential.

The jurisdiction project did not produce a portable legal remedy for every Internet identifier problem. It produced a disciplined example of how a global governance debate narrows when it reaches law. That example is valuable beyond ICANN names. It tells number-resource reformers to specify rights early: not "less jurisdictional capture" as a slogan, but which holder can move, contest, correct, transfer, restore, preserve, or obtain reasons under which instrument and against which institution.

The unresolved legitimacy question

After all the legal narrowing, a legitimacy question remains. Is it enough for a global Internet coordinator to be a California corporation with global commitments, community powers, mission limits, review mechanisms and targeted jurisdiction repairs? Reasonable people can disagree. Some will say yes because the alternative could be governmental capture, treaty paralysis or weaker private enforceability. Others will say no because global reliance should not depend so heavily on one domestic legal system, especially one capable of exporting sanctions effects and litigation costs.

The jurisdiction project did not decide that philosophical question. It made the practical settlement more tolerable for some parties. It reduced selected frictions. It recognized that contract choices and sanctions licensing are real governance problems. It left the larger constitutional objection mostly unanswered. That unresolved status should be stated plainly.

There is no shame in limited reform if the limitation is transparent. The danger is pretending that consultation volume equals settlement. A long, inclusive debate can still leave a party without a remedy. A short clause can create more legal protection than a hundred meeting hours. Legitimacy in a technical governance institution is not only about voice. It is about voice attached to a route of effect.

For ICANN, the route of effect after Work Stream 2 jurisdiction was selective. Applicants from sanctioned countries gained more concrete expectations. Contracted parties gained some law and venue design movement. Non-United States registrars received clarification. The wider public gained a record showing that jurisdiction concerns were heard. The corporation remained where it was. That is the result.

The next time ICANN opens a large accountability conversation, the jurisdiction project should serve as a warning label. Ask not only who spoke and how many pages were written. Ask which enforceable right changed. Ask which legal actor still holds the decisive power. Ask who can bring a claim, where, by when, and with what remedy. Ask which harms remain outside the changed instrument. If those questions are answered, participation can be honest even when it is limited. If they are avoided, participation becomes a way to decorate the existing legal order.

The jurisdiction project could not move the corporation because it was not designed to do so. Its value is narrower and more demanding: it shows that global governance reform must be measured at the point where words become rights. ICANN's California home survived the debate. The serious work is to make every remaining consequence of that home visible, bounded, reviewable and, where possible, less unfair to the people who never chose California but still depend on the systems coordinated there.

Sources reviewed