Summary

  • The first annual IGF meeting after permanent status is scheduled for Nairobi on 14-18 December 2026. As of 15 July, its final programme and outputs do not exist, so this is a live design audit with explicit later checkpoints rather than a retrospective verdict.
  • Permanent status removes the recurring mandate-expiry question and strengthens the case for stable staff, intersessional continuity and annual reporting. It does not change the Tunis boundary: the forum remains non-binding, has no oversight role and does not perform day-to-day technical operations.
  • The decisive 2026 test is whether every material recommendation names a capable recipient, states the requested act and evidence basis, sets a response date, labels its non-binding character, records receipt and later publishes adoption, rejection, referral, partial action or silence.
  • Public materials available by mid-July describe themes, calls, Policy Networks, programme architecture and ambitions for more actionable outcomes. They do not yet establish a common recommendation register or a duty for receiving institutions to respond.
  • The MAG should own output quality and classification; the Secretariat should maintain the public register; session organizers should validate the record; recipients should control substantive acceptance; and an independent reviewer should test completeness without pretending to enforce the recommendations.
  • Number-resource issues expose the boundary clearly. IGF discussion can identify evidence and refer concerns, while registries, ICANN, standards bodies, courts, public authorities and operators retain their distinct authority to decide and execute.
  • A Number Resource Society can contribute as a future-facing evidence and operator-rights institution only by making its own capacity and limits explicit. It should accept or reject referrals publicly rather than borrow authority from the prestige of a permanent UN forum.

The meeting has not happened, so the audit begins with time

The phrase “first permanent IGF” can mislead in two directions. It can suggest that the forum itself is new, although annual meetings have taken place since 2006. It can also suggest that the first meeting under permanent status has already produced an evidentiary record. It has not. The official IGF 2026 page schedules the twenty-first annual meeting for 14-18 December in Nairobi under the theme “Governing the Internet in the Age of Intelligence: Our Shared Responsibility.” This article is dated 15 July.

That timing matters because accountability begins with refusing to grade events that have not occurred. Session selection was still ahead: the official calendar opened calls for sessions, village booths and remote hubs from 29 June through 31 July, placed evaluations in August, anticipated a second open consultation in September and scheduled the programme later in the year. Draft messages are due on the final meeting day, while final messages, intersessional outputs and the summary report are scheduled for 15 January 2027.

The present evidence can support a design finding, not an outcome finding. We can examine the mandate, the MAG's stated architecture, the public timetable, the selected intersessional activities and the announced approach to outputs. We cannot claim that recipients answered recommendations that have not yet been issued. We cannot count completed actions, assess final wording or infer silence from a response period that has not begun.

This disciplined chronology is more than a caveat. It establishes the first line of the accountability record. On 15 July the design either contains the fields needed for later follow-up or it does not. On 18 December draft messages can be tested. On 15 January final outputs can be compared with the drafts. Ninety and 180 days later, recipient responses can be measured. A permanent forum should make that longitudinal inquiry easier than a temporary one.

Permanence changed continuity, not the verbs of authority

The UN General Assembly's Resolution 80/173, adopted on 17 December 2025, made the IGF a permanent United Nations forum. It continued the Secretariat hosted by the Department of Economic and Social Affairs, called for a stable and sustainable basis with appropriate staffing and resources, reinforced year-round activity and directed that outcomes be reported and transmitted into relevant UN discussions. These are significant institutional changes.

They answer a recurring weakness. A body awaiting another mandate extension has reason to spend political energy proving that it should continue. Permanent status permits longer staff planning, durable archives, recurring evaluation and commitments that span annual meetings. It also increases the responsibility to preserve institutional memory. A forum that expects to exist indefinitely can no longer excuse lost recommendations as an unavoidable consequence of episodic convening.

Yet the resolution also reaffirms the WSIS outcomes that define what the IGF is. The Tunis Agenda asks the forum to discuss public-policy issues, facilitate discourse, exchange information, identify emerging issues, advise, make recommendations where appropriate, build capacity and publish proceedings. It expressly says the forum has no oversight function, does not replace existing arrangements, is non-binding and has no role in day-to-day technical operations.

Permanence therefore strengthens continuity of attention. It does not add an order, licence, sanction, allocation, adjudication or implementation power. That distinction protects the forum from an impossible promise. It can be influential because it convenes people who possess different forms of authority elsewhere. It cannot make those external powers its own by describing an output as actionable.

The accountability chain proposed here respects that boundary. It asks whether an informed recommendation reached the right institution and received a visible disposition. It does not treat referral as command or non-response as contempt. The aim is not to manufacture enforcement. It is to make influence observable enough that entities and recipients can be judged on what they actually did.

The old execution gap is a missing chain, not a lack of ideas

IGF meetings have rarely suffered from a shortage of proposals. Hundreds of sessions generate messages, reports, transcripts, presentations and intersessional publications. Entities can often point to a strong paragraph about access, rights, security, artificial intelligence, infrastructure or development. The weak point comes after the paragraph.

Who was asked to act? Was that institution represented? Did it have legal, contractual, technical or budgetary capacity to do what the paragraph implied? Was the recommendation directed to one recipient or to a category so broad that no organization could know it owned the next step? Did the recipient acknowledge the message? Did it accept the evidence, reject the request, refer it to another body or explain that the matter fell outside its powers? Was any date attached?

Without answers, an output can be cited repeatedly while its practical status remains unknowable. Supporters may claim impact because the same idea later appeared in a policy. Critics may claim irrelevance because no direct causal line is visible. The receiving institution may have acted for independent reasons. A session organizer may count dissemination as implementation. The same document becomes evidence of success and evidence of failure depending on the speaker.

This is the execution gap: not the absence of executive power, which the IGF was never given, but the absence of a disciplined handoff between recommendation and the bodies that can choose to act. The gap allows prestige to substitute for responsibility. A message addressed to “all stakeholders” sounds inclusive while assigning the next step to nobody.

Permanent status raises the cost of that ambiguity. If the forum now has durable staff, annual reporting and a stronger claim to year-round relevance, it should preserve the complete history of each material recommendation. Otherwise the institution acquires permanence while its outputs remain disposable.

The public 2026 design is stronger on architecture than assignment

The first open consultations and MAG meeting, held in Nairobi and online on 24-26 June, supplied the clearest public view before this article's date. Its published meeting page identified goals that included coherence across annual, intersessional and national or regional activities, capacity development, contribution to WSIS+20 and Global Digital Compact implementation, possible Policy Labs and more actionable, policy-relevant outcomes.

The MAG also selected four Policy Networks after reviewing more than sixty proposals for intersessional activity. The annual site describes them as year-round multistakeholder platforms for dialogue, collaboration and output development. That choice can improve continuity because a standing group can refine an issue before Nairobi and remain available after the meeting.

The calendar is unusually useful. It states when proposals are submitted, evaluated and selected; when programme design is expected; when the meeting occurs; and when different outputs are due. Such publication creates a temporal backbone on which accountability could be built. The 2026 MAG membership and the roles of the Chair, host-country co-chair, Secretariat and Leadership Panel are also publicly described.

But architecture is not assignment. None of those materials, as visible by mid-July, establishes a universal output record with a named recipient, requested act, response date and disposition. “Actionable” is an aspiration, not a data structure for responsibility. “Contribution to implementation” says why an activity matters, not who must receive its result. “Policy Lab” describes a session form, not a commitment by any policy authority to answer.

The design finding is therefore mixed. IGF 2026 has the calendar, people and year-round forums needed to support a stronger chain. The common rule connecting those elements is not yet public. The opportunity remains open because session requirements and output practices can still be refined before December.

A recommendation needs seven fields before it needs more rhetoric

The minimum unit of accountability should be a recommendation record, not an entire meeting report. Each material recommendation needs seven fields.

First, it needs a recipient with a verifiable institutional name. A ministry, regulator, standards body, registry, company board, technical operator or UN body may qualify. “The community” does not. A recommendation may have several recipients, but each should have a distinct requested act.

Second, it needs a bounded action. “Improve trust” is a goal. “Publish a quarterly account of service restoration time using these measures” is an action capable of acceptance, modification or rejection. The field should state whether the request concerns research, consultation, rulemaking, procurement, technical deployment, disclosure, funding or coordination.

Third, it needs an authority note. The output should explain why the recipient can act and identify any limitation. A standards organization can consider a technical specification but cannot compel network deployment. A registry can alter its service within governing rules but cannot determine a criminal offence. A government can legislate within jurisdiction but cannot allocate global routing acceptance by decree.

Fourth, it needs an evidence basis. The record should link the session report, supporting study, dissenting evidence and material uncertainty. The IGF need not certify every claim as true. It should disclose what the recommendation rests upon.

Fifth, it needs an explicit non-binding label. This prevents a referral from being cited later as a UN directive. The label should travel with extracts and translations, not appear only in general terms elsewhere on the site.

Sixth, it needs a requested response date proportionate to the act. Acknowledgement might be due in thirty days; a substantive position in ninety or 180. The recipient remains free not to answer, but silence then becomes a dated fact rather than a vague impression.

Seventh, it needs a disposition: accepted, partly accepted, rejected, referred, already underway, outside authority, awaiting evidence, no response or superseded. A short explanation and link to the recipient's action complete the record.

Named recipients are a discipline against category error

Requiring a named recipient forces drafters to confront whether anyone actually possesses the imagined power. Internet governance language often compresses radically different institutions into “decision makers,” “the technical community,” “platforms,” “governments” or “industry.” Those labels are too coarse for responsibility.

Consider a recommendation that route-origin validation should increase. One recipient might be an RIR able to improve hosted RPKI availability. Another might be a standards body considering protocol work. A network operator decides whether and how validated information enters local route policy. A public authority may set procurement or critical-infrastructure expectations within law. A research group can measure adoption. These are connected actions, not interchangeable mandates.

The same discipline applies to artificial intelligence, child safety, access, encryption and data governance. A legislature cannot repair a vendor's security configuration directly. A company cannot resolve a constitutional rights question for a state. A conference host cannot commit every government present. Naming the recipient exposes the action chain and any missing link.

It also improves fairness. An institution should not be criticized for ignoring a recommendation it never received or could not lawfully execute. Conversely, a powerful recipient should not disappear inside a broad category after sending a representative to a high-profile session. The record asks for acknowledgement without claiming obedience.

Recipient naming will reduce the number of recommendations. That is a benefit. A concise set of well-routed requests has more evaluative value than hundreds of unowned aspirations. Session reports can still preserve wider discussion and minority views. The recommendation register should be selective enough that each entry can be followed.

The response date makes non-binding influence testable

Some defenders of open dialogue may resist deadlines because the IGF cannot compel a response. That objection confuses a requested date with an enforceable order. Libraries set return dates for voluntary loans; consultations set comment periods without forcing anyone to comment. A response date establishes the point at which the public record may fairly say that no answer has been received.

The date should match the recipient and action. A standards proposal may require a longer technical cycle. An emergency disclosure request may need days. A recommendation asking a public body to explain existing authority can reasonably receive an acknowledgement before a recommendation asking it to secure new legislation. The record should permit the recipient to state its own expected decision date.

Dates protect recipients as well as the forum. They prevent premature claims that an institution ignored a message. They distinguish a request under consideration from one abandoned. They also reveal whether the IGF releases outputs so late that relevance expires before the recipient can act. The official 2026 calendar's plan to finalize principal outputs by 15 January 2027 is helpful precisely because it creates a stable starting point.

Follow-up intervals should be fixed in advance. At thirty days the Secretariat records receipt. At ninety days it records the initial disposition. At 180 days it records substantive action or the recipient's schedule. At one year it notes closure, continuation or supersession. These intervals are observations, not sanctions.

A recipient that rejects a recommendation with evidence contributes more to institutional learning than one that offers ceremonial support and no act. The register should reward candour by preserving reasons rather than ranking acceptance as the only success.

Non-binding status must appear on every portable output

The IGF's legal and institutional limits are well known to specialists, yet outputs travel beyond specialist context. A sentence from a message can enter a ministerial briefing, company presentation, advocacy paper or funding application. Once detached from its source, “the IGF recommends” may be heard as “the United Nations requires.”

Every recommendation record should therefore contain a visible statement: this is a non-binding output of discussion and does not represent a negotiated decision, legal obligation, technical standard or direction to the recipient. Where a session had substantial disagreement, the record should say whether the text is a synthesis, a view expressed by many entities or a recommendation of a named intersessional group.

This label does not weaken the message. It clarifies the basis on which it deserves attention: evidence, participation, reasoning and relevance, rather than borrowed coercive force. Recipients can then accept a strong argument without implying that they recognize a new superior authority.

The distinction is especially important after permanence. Institutional longevity can be mistaken for hierarchical elevation. Repeating “permanent UN forum” beside an output may give it a status the underlying resolution did not confer. A portable non-binding label prevents that inflation.

The same rule should apply in all official translations and machine-readable feeds. A disclaimer hidden in an English methodology page will not accompany a translated excerpt. Accountability requires that the character of the output survive dissemination as faithfully as its substance.

Receipt is the first observable act, not proof of acceptance

The Secretariat should send each validated recommendation to a designated institutional address and publish the date, method and recipient office. Receipt can be confirmed by the recipient, by a public correspondence system or by another reliable record. A social-media mention, conference photograph or presence of a speaker should not count.

Acknowledgement means only that the recommendation entered the institution. It does not establish support, authority or implementation. This modest definition avoids a familiar inflation in which attendance is cited as endorsement. A minister who spoke on a panel may not have accepted a request. An engineer from a registry may not bind its board. A company entity may have attended personally.

Where no proper address exists, that fact is informative. The output may have named a category rather than an institution, or the institution may lack a public accountability channel. The Secretariat can ask the organizer to correct the recipient field before the item enters the register.

Recipients should be able to designate the correct office, split a recommendation among units or identify another body with stronger authority. Referrals should preserve the original date and reason. The chain then shows movement rather than silently replacing the addressee.

Proof of receipt is easy compared with proof of effect, but it is indispensable. Without it, later follow-up rests on an assumption. A permanent forum can afford the administrative discipline to establish this first fact consistently.

Disposition is more informative than a binary success claim

Implementation is rarely yes or no. A recipient may accept the goal while rejecting the proposed means. It may already be doing the requested act for reasons independent of the IGF. It may lack authority but refer the matter. A legislature may adopt part of a recommendation after amendment. A technical body may decide that evidence is limited public evidence. A registry may make an operational change while refusing a broader governance claim.

The public record should preserve those distinctions. “Accepted” requires a statement by the recipient and an identifiable next act. “Partly accepted” identifies the accepted and rejected elements. “Already underway” requires a pre-existing record and should not be counted as caused by the forum. “Referred” names the new recipient. “Outside authority” explains the boundary. “No response” records only silence by the observation date.

Claims of effect need stronger evidence. A later policy should cite the IGF output, or a responsible decision maker should confirm that it materially informed the decision. Temporal sequence alone is inadequate. Similar ideas often circulate through several venues, and the IGF may be one contributor among many.

This classification makes evaluation less flattering and more credible. Some outputs will show little external movement. Others will reveal that a modest research request produced a useful response. Over time the forum can learn which session forms, evidence practices and recipient relationships generate substantive engagement.

The purpose is not a league table of compliant institutions. It is a memory of what was asked, answered and changed, preserving rejection and uncertainty alongside adoption.

Responsibility inside the IGF must also be divided

The MAG should be responsible for the common output standard because it guides programme development and evaluates session proposals. It can require organizers seeking recommendation status to use the seven fields and can decline to classify vague aspirations as recommendations. This is quality control over IGF publication, not control over external recipients.

Session organizers should draft and validate the record. They know the discussion, evidence and disagreements. They should identify the recipient before the session where possible and invite a person able to explain institutional capacity. After the session they should confirm that the summary does not erase material dissent.

The Secretariat should maintain the public register, transmit records, log receipt and append recipient responses. Its role is custodial. It should not rewrite a rejection as progress or decide whether an external institution complied with its own law.

The Leadership Panel can promote dissemination and encourage responses, but it should not privately convert recommendations into commitments. Its interventions should be recorded when they concern a listed item. Prestige should help a message reach the right office, not obscure how it arrived.

An independent reviewer should sample records for completeness, recipient capability, evidence links and accurate disposition. The reviewer does not decide the merits of every digital-policy question. It tests whether the chain can be reconstructed and whether success claims match the record.

Finally, recipients own acceptance and execution. No internal IGF body should mark a recommendation implemented merely because an organizer reports activity. The institution that acted should identify the decision and evidence, while affected people remain free to contest its adequacy.

The MAG's programme power is real and should be auditable

The MAG does not govern the Internet, but it exercises consequential selection power over the annual meeting. It helps choose themes, session forms and proposals; integrates intersessional bodies and other tracks; and shapes which issues receive scarce room, translation, streaming and summary attention. That is enough power to require a visible rationale.

The 2026 materials publish evaluation stages and discuss predetermined criteria. The final record should show which proposals were received, found eligible, evaluated, selected, merged or declined, subject to privacy protection for individuals. It should disclose conflicts and recusals. A recommendation register cannot compensate for a programme whose selection is impossible to examine.

The MAG should also test recipient realism during evaluation. If a proposal promises an “action plan” but names no capable institution, evaluators should ask for correction. If a session concerns a body that has not been invited or refuses the premise, that absence should be disclosed. Proposals should not earn points merely for listing senior names.

Programme diversity and execution discipline are compatible. Underrepresented entities may be especially harmed by vague outputs because they invest scarce time and travel in a forum that later cannot show where their evidence went. A clear handoff gives contributions a trace beyond the room.

MAG accountability is therefore the first internal link. It cannot guarantee external action, but it can ensure that selected sessions capable of producing recommendations understand the minimum record expected of them.

Intersessional bodies need mandate and recipient labels of their own

Policy Networks, Best Practice Forums and Dynamic Coalitions vary in formation, participation, duration and editorial control. Their publications can be valuable, but the IGF umbrella does not make every statement equally representative. A permanent forum should make those differences easier to see.

Each intersessional output should identify who convened it, who drafted it, how comments were handled, whether a decision rule was used, what disagreement remains and whether the MAG approved the activity or the substance. It should distinguish collected practice from proposed policy. The recipient and non-binding fields then apply to any recommendation it contains.

This avoids institutional self-citation. A group should not publish a recommendation, cite it at the annual meeting and then treat its inclusion in final messages as independent confirmation. The evidence chain must identify common authorship and repeated reliance on the same underlying material.

Year-round activity can improve evidence. A Policy Network can ask a recipient for an early response, test feasibility and revise a request before Nairobi. It can also report that a proposed act falls outside the recipient's authority. Such correction is a sign of useful engagement, not failure.

The strongest 2026 opportunity lies here. The four selected Policy Networks can demonstrate the accountability chain before the annual meeting. If their outputs arrive with recipients, dates and visible limitations, the annual messages can inherit tested records rather than produce hundreds of new unowned statements in December.

National and regional initiatives are contributors, not a delegated hierarchy

The IGF ecosystem includes more than 175 national, regional and youth initiatives according to official descriptions. Their local knowledge can reveal implementation conditions that a global meeting misses. They do not, however, form subordinate branches that the global forum can direct, nor does use of the IGF name prove identical governance or independence.

When an NRI contributes a recommendation, the record should identify the actual local body, its participation basis and the recipient relevant to its jurisdiction. A global summary may aggregate recurring themes, but it should preserve differences in law, capacity and political risk. “NRIs recommend” is too broad unless a documented common act exists.

The reverse handoff matters too. A global output sent to an NRI should state whether the initiative is being asked to convene discussion, collect evidence or approach a public authority. It should not imply that the NRI can implement state policy or bind local operators.

Permanent status offers a chance to maintain these connections over several years. A recommendation could originate locally, reach a global body, return to a competent national recipient and later be reported with evidence. That chain would demonstrate the value of a distributed forum without pretending it is a federal institution.

The key is attribution. Every movement should preserve who said what, under which authority and with what response. Institutional coherence should mean traceability across difference, not flattening distinct bodies into one voice.

Number resources provide a hard test of the boundary

Internet number-resource governance makes vague assignment especially dangerous because registration, allocation, routing-security services and route selection involve different actors. RFC 7020 describes the Internet Numbers Registry System and the hierarchy connecting IANA, Regional Internet Registries and downstream services. RFC 6480 describes the RPKI architecture. Operators retain their own routing decisions.

An IGF session can examine registry continuity, route-security adoption, transfer friction, registration accuracy or operator rights. It can publish evidence and recommend that a named registry explain a service, that ICANN consider an accountability issue within its powers, that an IETF group examine a technical problem or that operators test a practice. It cannot allocate addresses, alter a registry entry, issue a route-origin authorization, decide corporate control or compel route acceptance.

The 2025-2026 revision of the RIR governance text illustrates the point. The NRO Number Council and ASO Address Council are developing the text through RIR and ICANN consultations. Existing RIRs, ICANN, communities and relevant legal bodies have distinct roles. An IGF message may illuminate the reform, but it does not adopt the document.

A recommendation register should therefore reject “the Internet community should secure routing” as unassignable. It should separate measurable acts: publish RPKI service recovery data, document a continuity test, study validator diversity, consider procurement expectations or report on local deployment. Each act goes to a body capable of answering.

This precision protects technical stability. It also prevents forum prestige from becoming a substitute for the consent and authority required by institutions whose decisions affect live network resources.

A live number-resource case shows the full chain

Imagine that an IGF 2026 session finds uneven public evidence about registry continuity during a governance crisis. The session hears from affected operators, a registry, technical experts and legal observers. Its strongest supported recommendation is that each RIR publish a comparable annual continuity exercise summary, including tested services, recovery objectives, material deficiencies and remediation dates without exposing sensitive security detail.

The recipient field names each RIR separately and the NRO for a common comparison method. The authority note says each RIR controls its own continuity operations, while common coordination requires agreement among the participating institutions. The evidence field links the session record, existing continuity obligations, operator testimony and dissent about cost. The non-binding label is explicit.

The request asks for acknowledgement within thirty days and a substantive position within 120. One RIR accepts and links an existing report. Another partly accepts but disputes the proposed comparison method. A third refers the common-method question to an NRO group. A fourth does not respond. The fifth says domestic security rules limit disclosure but offers aggregate measures. The NRO agrees to discuss a minimum format but does not promise adoption.

That outcome is not uniform implementation. It is nevertheless valuable. Operators can compare positions. The IGF can see where its evidence was persuasive and where legal or cost concerns remain. No response is recorded without being called a violation. No registry's acceptance is presented as commitment by all five.

At the next annual meeting, a session can examine the dispositions rather than restart the issue from memory. Permanence then produces cumulative inquiry. The forum has not become a registry supervisor; it has become a better custodian of evidence and institutional answers.

Public-sector continuity requires routes into budget and law

Many IGF recommendations are directed implicitly at governments, but “government” is rarely one recipient. A ministry may lead policy, a regulator may hold statutory power, a procurement office may set purchasing conditions, a parliament may legislate and an independent authority may enforce. Budget authority may sit elsewhere again.

The recipient record should identify the competent office and jurisdiction. Where several acts are required, it should split them. A request for national broadband measurement may require budget, statistical authority and regulator access. A recommendation on public-sector route security may involve procurement standards, network operators and incident reporting. One generic government response cannot stand for all of them.

The parliamentary and judiciary engagement tracks announced for IGF 2026 can improve institutional understanding, but attendance must not be equated with commitment. A judge cannot promise a future case outcome. A parliamentarian may advocate but not bind a legislature. A civil servant may explain policy without holding budget authority.

Public-sector continuity also extends beyond election cycles. A response should name the office rather than depend on the individual who attended. The Secretariat's register can preserve institutional contact changes and carry open items across years. That is a practical advantage of permanence.

Where political conditions make public response unsafe or impossible, the record should state the limitation and protect entities. Accountability must not expose vulnerable contributors. The recommendation can still identify a public institution while separating public evidence from confidential testimony.

Funding accountability belongs in the same chain

Resolution 80/173's call for stable and sustainable support creates an accountability issue inside the forum. Funding affects staff capacity, participation support, language access, intersessional work and the ability to maintain a recommendation register. It can also shape priorities indirectly.

The IGF should publish annual income by source category, restricted and unrestricted amounts, principal expenditure categories, travel-support distribution and any donor conditions, consistent with UN rules. It should explain which core functions remain unfunded and which outputs depend on temporary support. Host-country conference costs should be distinguished from Secretariat and year-round costs.

This does not mean that a donor bought a session whenever its interests overlap with a topic. It means the public should be able to test concentration and dependency rather than rely on reassurance. Stable funding is not neutral if a small set of institutions can determine which work continues.

The recommendation register itself needs protected core support. If follow-up depends on an interested sponsor or a short-term volunteer group, difficult dispositions may be neglected. Custody of the record should remain with the Secretariat, while independent review can be funded through a disclosed arrangement that prevents donor control over findings.

Permanence should make the funding history cumulative. A five-year series can reveal whether participation from developing countries, translation and follow-up improved after the resolution. That is a better measure than celebrating the word “sustainable” without showing what became sustainable.

Participation metrics should be linked to influence, not attendance alone

IGF reports often provide entity totals, country counts, stakeholder categories and online participation. These measures describe reach. They do not show who shaped the programme, drafted outputs, appeared as a named source or secured a response from a recipient.

The 2026 audit should add influence measures. How many accepted session proposals came from underrepresented regions? Who served as organizer and rapporteur? Which contributions entered final recommendations? Were dissenting views preserved? Which recipients answered requests originating from youth, civil society or small operators? Did travel support reach people with a substantive role rather than only attendance?

This inquiry must avoid reducing legitimacy to arithmetic. A large attendance does not authorize a recommendation, and a small expert group may produce strong evidence. The point is to detect whether formal openness coexists with concentrated editorial control.

Recommendation records can help because they identify authorship and evidence. A entity can see whether a contribution was incorporated, summarized inaccurately or omitted. Organizers can explain why evidence did not support a requested act. That is more respectful than counting the speaker and losing the substance.

Remote participation deserves the same treatment. The record should show whether remote interventions could alter text and whether technical failures excluded a constituency. Permanent status should support comparable measures across years, exposing persistent barriers that annual headline totals conceal.

An independent audit should verify the chain, not judge every policy

The IGF needs external scrutiny, but the audit scope must be bounded. An auditor cannot determine the correct global policy on every issue discussed. It can verify whether required fields exist, whether recipients were capable and contacted, whether links support claimed dispositions, whether non-binding labels travelled with outputs and whether reported effects are attributable.

A statistically meaningful sample can include main-session messages, Policy Network outputs, sessions with high-impact recommendations, items involving vulnerable groups and items the IGF cites as success. The reviewer should also inspect a random set to avoid selecting only prominent examples.

Findings should distinguish missing records from weak policy. A recommendation may be well documented and rejected. That is a complete chain, not an audit failure. Another may be popular but addressed to “all actors” with no route to action. That is a record failure even if the sentiment is admirable.

The reviewer should publish methods, conflicts and evidence limits. Sensitive entity information can remain protected. The MAG and Secretariat should answer findings and identify corrections. Repeated omissions should inform future session selection and training.

This audit strengthens institutional legitimacy without creating a supervisory body above recipients. Its authority concerns the accuracy of the IGF's own public claims. External institutions remain accountable through their own law, contracts and governing arrangements.

The 2026 scorecard should be frozen before the messages arrive

To prevent favorable reinterpretation, the first permanent-cycle scorecard should be published before December. It should measure design and results separately.

The design score asks whether the common recommendation format exists; whether organizers know it; whether recipient capability is checked; whether non-binding status is portable; whether the Secretariat will transmit and track items; whether response intervals are set; and whether an independent review is funded and appointed. These facts can be measured before the meeting.

The output score begins on 18 December. It counts material recommendations, the share with named recipients, bounded acts, authority notes, evidence links, uncertainty statements, non-binding labels and response dates. It should report medians and distributions, not one composite number that hides a weak field.

The follow-up score begins after 15 January 2027. It measures confirmed receipt, substantive responses, referrals, rejections, partial acceptance, action records, unsupported effect claims and unresolved silence at each interval. It should separate response from adoption and adoption from demonstrated effect.

The equity score examines authorship, evidence use and response rates across region, language, stakeholder category and participation mode, subject to privacy and sample limitations. The integrity score tests corrections, version history and whether dissent survives from session report to final message.

Freezing these measures now makes the audit reproducible. If IGF 2026 performs well, the record will show why. If it does not, permanent status provides another cycle in which to repair the design rather than another opportunity to redefine success.

What would count as a genuine first-year improvement

Success should be demanding but realistic. The IGF cannot guarantee that sovereign states, private firms or technical institutions implement its recommendations. It can guarantee the quality and traceability of its own handoffs.

A genuine improvement would be a public register launched before or immediately after Nairobi. Every main message framed as a recommendation would include the seven fields. At least one intersessional body would demonstrate the full chain from pre-meeting evidence to a recipient response. The final summary would distinguish recommendations, observations and contested views. Corrections and translations would preserve status labels.

The Secretariat would transmit records by a published date and record acknowledgement. The MAG would issue a short account of how output discipline affected programme decisions. The Leadership Panel would report outreach tied to listed items. An independent reviewer would publish an initial completeness assessment before the next annual call.

Most importantly, the IGF would refrain from overstating impact. It would identify rejection, non-response and independent parallel action honestly. A small number of traceable recommendations would count as more progress than a large collection of messages with no owner.

The first year need not solve attribution perfectly. It should establish the habit that every claimed recommendation has an address and every claimed result has evidence. That habit is the institutional dividend permanence can deliver immediately.

What would show that the execution gap survived unchanged

Failure would not be the absence of universal implementation. It would be the reproduction of old ambiguity under a new permanent label.

Warning signs include final messages directed to “stakeholders” without institutional names; action language without an authority note; session reports published after recipient decisions have moved on; outcome pages that count downloads or mentions as impact; and annual summaries that list themes without dispositions. Another warning is a private follow-up effort described publicly only through success stories.

If intersessional groups cite their own earlier documents as independent confirmation, the evidence base remains circular. If the Leadership Panel announces influence without identifying recipient acts, prestige has replaced attribution. If national or regional initiatives are treated as endorsing global text merely because they share a name, coherence has become appropriation.

The most serious failure would be authority inflation: a recipient or advocate presenting a non-binding message as a UN mandate. The Secretariat should correct such use publicly while remaining neutral on the recommendation's merits. Permanence increases the risk because the institutional description sounds stronger.

These failure signals are observable. Naming them before Nairobi prevents a later debate from turning entirely on impressions.

NRS can model a cleaner future handoff without claiming succession

The Number Resource Society represents a positive future direction when it focuses on operator rights, accurate registration, evidence, portability and replaceable institutions rather than territorial prestige. That direction is relevant to the IGF accountability problem because NRS can demonstrate how a recipient answers without pretending that forum attention grants authority.

If an IGF output asks NRS to study registration error, continuity or route-security evidence, NRS should publish a receipt, state its current capacity, accept or decline the task and identify what it can deliver. If it lacks operational authority, it should say so. If it produces a comparison or test method, the result remains evidence offered to operators and institutions, not a command.

NRS should apply the same standard to its own recommendations: named recipients, bounded requests, dates, evidence, uncertainty and visible dispositions. It should not cite an IGF appearance as recognition or treat participation as endorsement. A future institution earns confidence through demonstrable service and accountable authorization, not by collecting institutional associations.

This restraint is strategically strong. It separates a future registry direction from the recurring habit of converting convening prestige into power. Operators can evaluate the quality of the evidence and the usefulness of the proposed rights without being asked to accept a fabricated mandate.

The relationship can therefore be complementary. The IGF broadens discussion and routes evidence. NRS develops operator-centred proposals and, where validly authorized, services. Existing registries, ICANN, standards bodies, courts, public authorities and networks retain their distinct functions. A visible handoff preserves cooperation without institutional merger.

The next checkpoints are already knowable

This real-time audit should be updated against fixed dates, not rhetorical milestones. On 31 July the session call closes, allowing inspection of proposal requirements. At the end of August, evaluation practice can be examined. The September consultations can show whether a common output record is adopted. By 18 September the schedule is expected to take shape, and by 19 October main sessions are due to be finalized.

During the Nairobi meeting, draft messages can be tested for recipient and status fields. On 8 January 2027 session reports are due. On 15 January final messages, intersessional outputs and the summary report are scheduled. Thirty, ninety and 180 days after final publication provide the first receipt and disposition observations.

At each checkpoint, the question remains narrow: did the institution add a link in the chain? A changed theme or another impact statement does not answer it. A published field, transmission record or recipient response does.

The audit record should preserve versions. If a recipient is added after criticism, that is a useful correction and should remain visible. If draft uncertainty disappears from final text, the change deserves explanation. If a recommendation is withdrawn, the reason should be recorded rather than the entry deleted.

Permanence gives the IGF enough time to learn from this evidence. It also removes the excuse that the annual cycle ended before anyone could follow the result.

Conclusion: permanence needs a memory of responsibility

The first permanent IGF has an opportunity that earlier cycles did not possess in the same form. Its future is no longer conditioned on another ten-year renewal. The UN has called for stable support, stronger intersessional activity, wider participation and clearer contribution to other public discussions. The 2026 calendar supplies identifiable stages from thematic input to final output.

But the execution gap will remain until the forum records where each serious recommendation goes. Named recipient, bounded act, authority note, evidence, non-binding status, response date and disposition are modest requirements. They do not turn discussion into enforcement. They turn institutional memory into accountability.

As of 15 July, the public design does not yet demonstrate that common chain. That is a provisional finding, not a prediction of failure. The annual meeting is five months away, the final programme is not set and the output period extends into January. The remaining design window should be used.

If IGF 2026 publishes a small set of traceable recommendations and follows them honestly through rejection, referral, silence and action, permanence will have produced a meaningful first reform. If it produces familiar messages addressed to everyone and owned by nobody, the institution will be more durable while its accountability remains episodic.

The choice is not between a powerless forum and a global executive. It is between untraceable influence and disciplined, non-binding handoff. The latter fits the IGF's mandate, respects the authority of receiving institutions and gives future entities evidence that their contributions travelled farther than the conference room.

Sources