Summary
- The MAG is not elected by the Internet's users, companies, governments or technical operators. The United Nations Secretary-General appoints its members after an open nomination process administered through the UN Department of Economic and Social Affairs and the IGF Secretariat.
- Any interested individual or organisation may nominate a candidate, and self-nomination is allowed. Stakeholder recommendations can inform selection, but the public record does not consistently show a complete nominee-to-nominator-to-appointment trail or the comparative reason for each choice.
- Members serve in their personal capacity while being assigned to stakeholder categories and expected to maintain community links. This protects individual judgment but makes "representation" ambiguous unless consultation and reporting duties are specified.
- Current terms are one year, generally renewable for two further consecutive years after performance evaluation. Low participation may prevent automatic renewal, but the published terms do not set out an equally clear procedure for removal during a term, notice, response, vacancy filling or review.
- Geographic, gender and stakeholder balance are valuable composition controls. They should be accompanied by public tenure, nomination, affiliation, recusal, evaluation and exception records so diversity is not mistaken for delegated consent.
A diverse table is not a chain of authorisation
The MAG performs work that looks modest and carries substantial agenda influence. It advises the Secretary-General on the programme and schedule of the annual IGF. It identifies themes, reviews proposed sessions, helps shape main sessions, connects intersessional activity and supports outreach. It does not make law, allocate Internet resources or supervise technical operations. Yet it decides which global digital-policy questions receive scarce common attention and how those questions are framed.
That role makes composition important. A programme group drawn only from one government bloc, one industry or one region would predictably miss problems. The MAG therefore seeks balance across governments, civil society, the private sector and the technical community, alongside geography and gender. Current materials describe roughly 40 members, with government holding about 40 per cent and the other three groups sharing the remainder more or less equally.
Balance is an answer to concentration. It is not by itself an answer to delegation. A civil-society label does not show which organisations or publics selected the person. A private-sector label does not establish authority from companies of different sizes and regions. A technical-community label does not show consent from network operators, engineers or standards contributors. A government official may have a clear state appointment while lacking authority to speak for other governments in the region.
The distinction is easy to lose because a well-balanced roster looks representative. Photographs and biographies display regions and sectors. The programme can plausibly be called multistakeholder because people with different institutional locations participate. But representativeness as a property of composition is different from representation as a relationship of authorisation, instruction, reporting and correction.
The MAG does not need to become a global parliament. It does need to tell the truth about its authority. Members are appointed advisers serving personally. Their experience, networks and consultation can make their advice excellent. The legitimacy claim should rest on an open appointment chain, transparent reasons and accountable conduct, not on an implication that the world's stakeholders elected a miniature version of themselves.
The founding mandate placed the closing authority with the Secretary-General
The Tunis Agenda asked the United Nations Secretary-General, through an open and inclusive process, to convene a new forum for multistakeholder policy dialogue. It did not create an elected council or specify a nomination college for the programme advisers. The legal and institutional line therefore began with the Secretary-General's convening responsibility.
On 17 May 2006, Secretary-General Kofi Annan established the first Advisory Group to help prepare the inaugural meeting in Athens. The announcement listed 47 people including chair Nitin Desai. The associated group breakdown is sometimes reported as 46 members because it excludes the chair. This numerical difference is mundane but instructive: even basic roster statistics require a stated denominator.
The members came from governments, business and civil society, including academic and technical communities, and from all regions. The Secretary-General appointed the chair, who could choose special advisers. The group met after an open consultation and was tasked with preparing the agenda and programme. Its recommendations went to the Secretary-General.
Names had been invited through governments and stakeholder networks. The edited institutional history Internet Governance Forum: The First Two Years says names were collected, collated and sent to the Secretary-General for selection. It reports a composition of 22 government, six business, seven civil-society, ten Internet-community and one media member, with several people having overlapping affiliations and some non-governmental candidates proposed by governments.
This origin established the basic chain that persists: communities and states put forward names; an administrative centre compiles and evaluates; the Secretary-General appoints. The chain was multistakeholder in its sources and composition, but the final authority was singular. No stakeholder group could demand that its preferred nominee be seated.
The arrangement was defensible for a body advising the Secretary-General's forum. It should never be described as if the stakeholder groups jointly appointed their delegates. They supplied candidates and recommendations into a UN appointment process.
The modern route is open at the nomination gate
The current MAG renewal FAQ makes the entry route broader than many readers may assume. On behalf of the Secretary-General, the Under-Secretary-General for Economic and Social Affairs issues a public call. Any interested individual or organisation may nominate a candidate, and individuals may nominate themselves. The form identifies the stakeholder group with which the candidate is associated.
This openness reduces the power of established gatekeepers. A candidate need not wait for a global civil-society network, trade association or technical organisation to issue an endorsement. A newer entity from an underrepresented country can enter the formal field directly. The possibility of self-nomination is especially important where no organisation has a legitimate claim to select for a diffuse constituency.
Stakeholder organisations and networks still matter. They circulate calls, identify experienced entities, compare candidates and submit recommendations. Their endorsement can supply evidence of trust and community linkage. It can also reproduce insider networks if the recommending group is small, poorly advertised or mistaken for the entire stakeholder category.
The FAQ says candidates are assessed using factors including breadth of support, likely contribution, knowledge, expertise and the diversity of viewpoint they would add. Composition decisions also consider region, stakeholder group, gender, developing-country participation and institutional variety. Current guidance gives countries that have been absent or underrepresented an advantage, subject to other requirements.
These are sensible factors. They are also capable of conflict. Breadth of support may favour candidates attached to large organisations. Expertise may favour long-serving insiders. Bringing in newcomers may reduce immediate institutional knowledge. Geographic correction can compete with stakeholder balance. Gender parity can be achieved while class, language, disability or employer concentration remains invisible.
An open gate does not make the comparison self-explanatory. The public needs to know how criteria were applied in the aggregate, which tradeoffs governed a particular year and whether the published call matched the eventual composition. Privacy may justify withholding personal references and unsuccessful applications. It does not justify keeping the architecture of selection obscure.
Nomination is evidence of support, not a transferable mandate
The word "nomination" can conceal several different acts. A ministry may nominate an official under a formal instruction. A trade association may nominate after a member process. A civil-society network may use an open call and a panel. A technical list may reach rough agreement among active subscribers. An employer may put forward a staff member. A person may nominate themselves.
These routes do not create the same authority. Formal state nomination can show a clear chain from a government, but the official still joins the MAG personally and does not receive a vote weighted by population. An association's nominee may be accountable to its members but not to every company in the private sector. An open network may be accessible while reaching only people already connected to its channels. Self-nomination demonstrates willingness, not external endorsement.
The current FAQ contains a revealing duality. It says selected candidates serve in their personal capacity, while also saying they represent the interests of an entire stakeholder group. The MAG terms of reference use more careful language: members serve personally but are expected to have established linkages with their respective stakeholder groups.
"Linkage" is the more defensible concept. A member can consult relevant communities, bring evidence, explain decisions and invite correction without claiming to have been elected by all of civil society, business or the technical community. The quality of the link can be evaluated through conduct.
Calling a member the representative of an entire group creates an impossible burden. There is no global register of civil society, no universal electorate of Internet businesses and no single technical constituency. Interests within each category conflict. Large platforms and small providers are both private sector. Human-rights organisations and community-development groups are both civil society. Protocol designers, cybersecurity responders and access-network operators may all be technical community while disagreeing sharply.
The appointment record should therefore preserve the exact nomination route. It should say whether the person was self-nominated, organisation-nominated or recommended through a documented community process; identify the nominator with consent; and avoid upgrading endorsement into constituency-wide delegation. That clarity protects nominees from claims they could never satisfy.
The administrative middle deserves more attention than the ceremonial announcement
Public descriptions often jump from community nominations to the Secretary-General's appointment. Between them sit the IGF Secretariat, UNDESA officials, consultation with stakeholder channels and the practical task of producing a balanced slate. This administrative middle is where most comparisons and tradeoffs must occur.
The Secretariat knows the Forum's workload, attendance patterns and community history. It can identify whether a candidate has organised sessions, participated constructively, consulted others or has the time to work. UNDESA connects the process to the Secretary-General and applies broader UN composition expectations. Stakeholder recommendations supply external knowledge. The Secretary-General's office supplies final institutional authority.
No one should infer that the Secretary-General personally reads every application and calculates every balance. Nor should an appointment press release be treated as evidence that no administrative recommendations shaped the result. Large institutions act through staff. Legitimacy requires a truthful description of those roles.
Independent research has challenged the opacity of this middle. The Centre for Internet and Society's 2016 Mapping MAG study argued that early nomination records and criteria were weak, that roster archives had gaps and that Secretariat recommendations and New York-based selection interacted in ways entities could not reconstruct. Some of its claims rely on interviews and the author's interpretation and should not be accepted as official findings. The study is still important evidence of what a serious external observer could not verify from published records.
The answer is not to expose confidential applicant assessments. It is to publish an annual selection statement: number of candidates by region and stakeholder group; routes of nomination; criteria and their priority; seats available; conflicts managed; reasons for any exceptional rule; and a composition table showing how the final slate answers the stated need. If a stakeholder group's recommendation is not followed, the statement can explain the category-level tradeoff without ranking named unsuccessful candidates.
Ceremonial appointment remains with the Secretary-General. Administrative accountability belongs to the whole chain.
Personal capacity protects judgment and weakens the fiction of instruction
Personal-capacity service has real value. A company employee should not be required to defend every corporate position. A civil servant should be able to consider evidence outside a national brief. A technical expert should not treat the seat as property of an employer. A civil-society entity should not become the instructed vote of the organisation that paid for travel.
This independence helps a programme committee find workable combinations across sectors. MAG members review sessions and design discussions, work that benefits from judgment rather than bloc voting. If every agenda choice had to be cleared by external principals, the Forum could reproduce the rigid statements of an intergovernmental negotiation.
Personal service does not erase institutional position. Employers provide salary, time, travel, information and professional networks. A member's understanding of risk is shaped by experience. Government officials may be accountable under public-service rules. Business representatives may have duties to employers. Advocates may have organisational missions. Disclosure, recusal and plural evidence are still necessary.
Nor does personal capacity answer whom a member should consult. The terms expect community linkage, but the form of linkage is not standardised. One member may run open calls, publish notes and report back. Another may rely on familiar contacts. Both appear under the same stakeholder label.
A light consultation statement would improve accountability without turning members into delegates. At the beginning of a term, each member could identify the communities and channels they expect to use, any limits on their reach and relevant affiliations. At the end, they could report consultations, issues raised, programme changes influenced and unresolved concerns. The report should not list private speakers or claim statistical representation.
This would shift the question from identity to practice. A person does not represent civil society because a biography says so. The person contributes a civil-society perspective credibly when they maintain open links, carry diverse evidence, disclose limits and answer back. The same applies to every non-governmental category.
One-year terms create an annual accountability point and an annual dependency
The current terms of reference state that members are appointed for one year. The term is generally renewable for two consecutive years, contingent on yearly evaluation of engagement and contribution. Roughly one third of the membership rotates annually to bring new viewpoints while preserving continuity.
Short terms create regular correction. A member who cannot attend or contribute does not need to occupy a seat indefinitely. Regions and communities that were absent can enter. The programme does not become the property of a permanent cadre.
The same design creates dependency. A first-year member who hopes to continue knows that performance will be assessed inside the appointment chain. If evaluation criteria are unclear, visible agreement can be safer than principled dissent. Members with the time and resources to attend every call may outperform people whose participation is constrained by time zones, care, disability, connectivity or unfunded employment, even when the latter bring rarer knowledge.
The FAQ says attendance, contribution and participation in IGF activities inform the end-of-term evaluation and that low performers may be excluded from automatic renewal. This is clearer than a purely discretionary extension. It still leaves questions. Who scores performance? Are members warned? Can they explain barriers or contest incorrect attendance data? How is intellectual contribution assessed without rewarding volume? Is dissent protected? Are community consultation and conflicts considered?
Automatic renewal language also deserves precision. The appointment is formally for one year, and the Secretary-General retains appointment prerogative. "Generally renewable" should not create an entitlement, while "automatic" should not hide a fresh exercise of public authority. Each continuation is both a tenure expectation and a renewed appointment.
An annual record should list the term number, evaluation status and outcome. It need not publish personnel detail. Categories such as renewed after satisfactory service, completed maximum terms, voluntarily departed, not renewed after documented low participation, or exceptional continuation would make rotation reviewable without humiliating individuals.
The published rules are clearer about non-renewal than mid-term removal
The question "Who can remove a MAG member?" does not have an equally direct public answer. The Secretary-General appoints members for one year. The published FAQ explains how low participation can prevent automatic renewal. The terms describe duties and yearly evaluation. They do not set out a complete mid-term removal process with grounds, initiator, investigation, notice, response, decision maker, vacancy procedure and review.
Institutionally, the appointing authority and UN administrative chain would have to address a serious case. A member may resign, become unable to serve, change roles, breach conduct expectations or cease participating. The 2026 guidance says annual turnover also reflects members informing the Forum that they can no longer serve. But inference from appointment power is not a substitute for a public rule.
The absence matters even for an advisory body. If removal is practically impossible until the next annual cycle, an inactive seat can leave a region or stakeholder group underrepresented during programme selection. If removal can occur informally, members may not know the standard protecting their independence. If the Secretariat can simply stop involving someone, formal appointment and actual participation can diverge.
A proportionate rule should distinguish four events. Voluntary resignation requires notice and a public effective date. Administrative vacancy covers death, incapacity or sustained inability to participate. Performance non-renewal occurs at term end under published criteria. Early removal addresses serious misconduct, undisclosed conflict, misuse of position or persistent non-performance after notice.
For early removal, the member should receive the allegation, a chance to respond and a reasoned decision from the proper authority. Urgent interim measures may be possible for safety or integrity. A replacement should normally come from the same region and stakeholder balance need, but not automatically from the same nominator. The public notice should protect sensitive details while stating the authority and ground category.
This is not bureaucratic excess. Clear removal rules protect both the Forum and the member. They make correction possible without turning annual renewal into the only disciplinary tool.
The chair has a separate appointment chain and additional framing power
The MAG chair is not simply the member who receives the most peer votes. The chair's terms of reference state that the Secretary-General appoints the chair for one year, subject to possible renewal. The chair guides the group, facilitates consensus, coordinates with the Secretariat and represents the MAG externally under stated procedures.
This role has more than ceremonial influence. A chair organises discussion, interprets emerging agreement, decides when to close an item, connects competing proposals and speaks for the group. Even without a formal casting vote, procedural judgment can shape the programme.
Separate appointment may provide stability and a clear link to the convenor. It also means a multistakeholder group does not wholly select its own presiding officer. The 2026 press release records that Secretary-General António Guterres appointed Jennifer Chung, from the technical community, as chair while renewing 40 members. Eighteen of the members were new.
The public should therefore see chair selection as its own track. A roster balance table cannot explain it. The annual notice should state the eligibility pool, consultations undertaken, term history, selection criteria, stakeholder and regional rotation considerations and whether MAG members were invited to advise. The Secretary-General may retain the final choice while making the recommendation path visible.
Chair accountability also requires a distinction between representing the MAG and speaking personally. External engagements should identify the approved position, consultation with members and any material disagreement. The chair's terms already require disclosure and UNDESA review for external representation. Publishing a compact engagement register would make that control operationally visible.
An independently appointed chair need not be an illegitimate chair. The legitimacy depends on accurate authority labels, a reviewable appointment and restraint in claiming collective support.
Composition categories are useful controls and crude descriptions
The modern MAG uses four principal stakeholder groups: government, civil society, private sector and technical community. Media appeared separately in earlier composition descriptions and academia has sometimes been grouped differently. Former host countries have standing participation, while treaty-based intergovernmental organisations may have observer seats.
Categories make balancing possible. Without them, a roster of impressive individuals could conceal that most work for governments or large technology companies. Regional groups and gender tracking add further checks. The 2026 guidance aims for about 40 per cent government participation and equal shares among the other three groups, balanced across regions by population and number of member states.
But categories are not natural facts. A university researcher may fit civil society, academia or technical community. A registry executive may be technical community or private sector. A state-owned operator complicates government and business. A consultant may have several clients. The first Advisory Group already contained overlapping identities, and its retrospective breakdown assigned each person to the primary group that proposed them.
Classification can therefore determine apparent balance. If the nominator chooses the label, candidates have an incentive to enter through the category with vacancies. If administrators reclassify, they exercise additional discretion. If affiliation changes during a three-year sequence, the original label can become stale.
The roster should show both the appointment category and current material affiliations. A change of employer should not automatically terminate service, because members serve personally. It should trigger an update and, where necessary, a conflict review. Aggregate reports should preserve the original category so selection history remains interpretable.
Balance within categories matters too. Ten private-sector seats occupied by multinational firms do not show the experience of small providers, local platforms or cooperatives. Civil society concentrated in globally funded organisations may miss local-language groups. Technical seats concentrated in naming institutions may miss access-network operations or security response. Regional nationality does not establish residence, work focus or connection to underserved communities.
The answer is not endless quotas. It is richer disclosure and a statement of which dimensions the selectors prioritised that year. Composition is a tool for reducing blind spots, not proof that blind spots have vanished.
Exceptional years expose where the real discretion sits
Routine rules become clearest when an institution departs from them. For 2025, the Forum used an exceptional approach tied to the twentieth anniversary, the WSIS+20 review and a shortened preparation period. The nomination pool was restricted to former MAG members and members of the original Working Group on Internet Governance. Forty people were appointed for one year without the usual prospect of immediate renewal.
The 2025 membership page openly describes the exception and its rationale: experience was prioritised to navigate the review and strengthen future working methods. That disclosure is valuable. It also reveals that the broad community nomination model is a policy choice within the Secretary-General's appointment authority, not an unchangeable constitutional requirement.
An alumni-only pool favoured institutional memory and excluded capable people without prior status. The tradeoff may have been reasonable under time pressure. It should be evaluated against the results and not quietly normalised. The narrow pool also demonstrates why "multistakeholder" cannot be inferred from open nomination in every year; 2025 remained compositionally multistakeholder while its candidate gate was intentionally restricted.
For 2026, guidance returned to a broader approach and used the 2024 composition as a continuity reference so eligible members did not lose the opportunity to complete up to three terms because of the exceptional year. The 2026 renewal announcement named 40 members, 18 new, and a new chair.
This repair was sensible but complex. A person can have a calendar-year gap while a tenure rule looks back to an earlier cohort. Without a public tenure register, outsiders cannot verify equal treatment. Exceptions create obligations to show the baseline, deviation, affected members, expiry and restoration path.
The Forum is now permanent following the 2025 UN review. Permanent status increases the importance of stable appointment rules. An institution can retain flexibility for exceptional years while requiring that every exception be explicit, time-limited and reviewed.
Geography should track voice, not only passports
Geographic balance corrects a persistent problem in global Internet institutions: people with resources to attend international meetings often come from a small set of countries and cities. Giving an advantage to unrepresented or underrepresented countries can bring new infrastructure conditions, legal systems, languages and user experiences into programme design.
Nationality alone is an imperfect proxy. A candidate from an underrepresented country may work for a global organisation in a major policy centre. A migrant or diaspora expert may understand several regions. A person may hold one passport, live in another country and work primarily on a third. Treating these lives as errors would be absurd, but counting only nationality can overstate the diversity of operational experience.
The selection form should distinguish nationality, residence, principal work region and the communities with which the candidate has active links, all with appropriate privacy choices. Public rosters can present only what is necessary. Aggregate selection analysis can show whether geographic correction reached beyond formal nationality.
Regional group balance also should not hide differences within regions. Africa is not one connectivity condition. Asia-Pacific contains enormous disparities in population, income, language and state power. The Western European and Others Group combines countries with very different policy roles. Latin America and the Caribbean and Eastern Europe contain their own concentration patterns.
Funding affects whether geographic appointment becomes meaningful participation. Members are unpaid, and some may receive travel support subject to available resources. A seat without time, connectivity, interpretation or travel can become symbolic. Performance evaluation must not punish resource constraints the institution failed to address.
The accountability loop is straightforward: selection creates a geographic expectation; support makes participation possible; consultation supplies regional evidence; annual reporting shows whether the expectation was met. The passport is the beginning of the inquiry, not its conclusion.
A traceable tenure ledger would make continuity and capture visible
The IGF website now publishes current and past member lists, affiliations, stakeholder groups and terms served for many recent years. The 2026 roster links previous lists back to 2011 and annual renewal announcements. This is substantial progress compared with the archival gaps identified in earlier research.
A reader still has to assemble the longitudinal story across pages. Names change format, affiliations change, observers and former hosts appear alongside appointed members, and chair terms may need separate interpretation. A consistent downloadable tenure ledger would allow the public to see appointment and influence patterns without manual reconstruction.
Each row should contain a stable public member identifier, name, appointment year, term number, stakeholder category at appointment, region basis, nominator type, appointing notice, current disclosed affiliation, chair or member role, renewal outcome, end reason and any exceptional rule. It should distinguish appointed members from host-country entities and intergovernmental observers.
The ledger would reveal both concentration and loss of memory. Repeated service by a small network may indicate expertise or capture. High turnover may indicate openness or an institution unable to retain knowledge. The data do not decide which interpretation is correct. They make the question testable.
It would also improve fairness. Candidates could verify whether term limits are applied consistently. Communities could see whether nominations lead to appointments over time. Administrators could detect duplicate or stale affiliations. Scholars could analyse composition without relying on private spreadsheets.
Privacy does not require publishing unsuccessful candidates indefinitely. The tenure ledger covers people who accepted public appointments. A separate annual nomination report can use aggregate counts and publish nominator names only with consent. The principle is proportional transparency: more detail for the exercise of public institutional authority, less for individuals who were considered but not selected.
Renewal should evaluate service without rewarding conformity
Attendance is easy to count. Contribution quality is harder. A member who speaks on every call may add less than one who identifies a missing constituency, corrects an assumption or organises a difficult compromise. A member who dissents from a popular programme choice may be performing the advisory role well.
Evaluation should combine objective and qualitative elements. Objective measures can include meeting participation, completed reviews, conflict disclosures, response to assigned work and consultation activity. Qualitative review can assess preparation, collaboration, evidence, respect, independence and contribution to the group's range of knowledge.
No single chair or staff member should make an unexplained continuation judgment. A small review process could include the chair, Secretariat and peer input, with conflict controls. The Secretary-General's appointment authority remains final, but the recommendation should follow a published standard.
Members should receive a midyear statement if performance is at risk. It should identify missed duties and available support. At term end, the member should be able to correct factual errors and explain barriers. The process need not become adversarial litigation. Basic notice improves both performance and fairness.
Dissent protection should be express. Non-renewal cannot be presumed retaliatory whenever a member disagrees, but opaque evaluation makes that suspicion difficult to answer. Recording that reasoned dissent and minority views are legitimate contributions would reduce pressure toward visible unanimity.
Community linkage should count, but not by popularity alone. A member who reports uncomfortable evidence may receive less endorsement from powerful organisations. The relevant question is whether consultation was credible and the evidence was carried faithfully, not whether every external stakeholder approved the final advice.
The annual appointment statement should report the number of eligible returning members, renewals, voluntary departures, completed terms and performance-based non-renewals in aggregate. Exceptional extensions should be named and reasoned because they alter the promised circulation of opportunity.
Stakeholder communities need duties when they claim a nomination role
The UN side is not the only possible source of opacity. A nomination described as "from the community" may have passed through an association board, an open mailing list, a small selection panel or one well-connected coordinator. The final appointment record cannot be more accountable than the recommendation it receives unless the route is described.
No single method should be imposed on every group. Governments have formal administrative systems. Companies may act through trade bodies or open calls. Civil society often depends on overlapping networks that resist central membership. Technical communities may prefer open participation and rough consensus to corporate voting. Diversity of method is consistent with multistakeholder practice.
Minimum evidence can still be common. A recommending body should publish the call, eligibility, selector identities and conflicts, criteria, consultation period, number of candidates and method used to reach the recommendation. It should state the constituency it can credibly reach and avoid claiming a monopoly over the whole category. If confidentiality protects candidates, the group can publish aggregate information and the selected names after consent.
The candidate should know what follows from the endorsement. Is there an expectation to consult the nominating network, submit reports or exercise independent judgment? Because MAG service is personal, no nominator should control a member's decisions or threaten withdrawal merely because the member disagrees. The proper accountability is explanation and evidence, not instruction.
Plural nomination routes can be a strength. A Secretary-General choosing among self-nominees, open-network recommendations and established organisations can correct the blind spots of each. But the selection statement should identify the mix. Otherwise, a well-organised intermediary may become a permanent unofficial gatekeeper while the formal process continues to be described as open.
Communities should also be able to report a broken link. If a member ceases all consultation, falsely claims endorsement or no longer has the disclosed relationship, the nominator should have a channel to submit evidence. That report should trigger review, not automatic removal. Appointment belongs to the Secretary-General, and personal capacity protects the member from recall by a private organisation.
This arrangement keeps each authority in its place. Communities recommend and supply evidence. The UN appointment chain selects and, where necessary, corrects. Members advise independently and report honestly. None can claim the powers of all three.
A practical appointment compact would clarify the whole chain
The MAG does not require a complicated electoral constitution. It needs a compact that connects existing practices and fills the visible gaps.
First, the annual call should state vacancies, stakeholder and regional needs, term rules, criteria, evidence requested, evaluation stages, expected dates and who performs each stage. Any exceptional candidate restriction should appear at the top, not after nominations begin.
Second, every nomination should record its route: self, organisation, government or documented community process. Candidates should consent to publication of selected information if appointed. Endorsements should be evidence, not votes, unless a nominating community has expressly defined them as such.
Third, the administrative assessment should produce an internal conflict-controlled record and a public aggregate report. The public report should explain tradeoffs and show the candidate pool without exposing private references.
Fourth, the Secretary-General's appointment notice should identify the authority, list members, distinguish the chair, state term numbers and link the applicable criteria and any exception. It should not imply that stakeholder categories elected their members.
Fifth, members should publish affiliation, conflict and consultation statements. Changes should be updated during the year. Recusal records should identify the subject and participation limit without revealing confidential commercial or personal information.
Sixth, annual evaluation should use published measures, notice and factual correction. Renewal, completion, resignation and non-renewal categories should update the tenure ledger.
Seventh, an early-vacancy and removal rule should identify grounds, interim steps, response rights, decision authority and replacement method. The rule should protect both programme continuity and member independence.
Finally, a limited process review should accept complaints about procedure, classification, undeclared conflict or inaccurate records. It should not promise every unsuccessful candidate a merits appeal. Corrections, explanations and future reform may be adequate remedies.
This compact would make the real model stronger: an appointed advisory group informed by open and diverse nomination, not an imaginary elected chamber.
The MAG's legitimacy depends on visible limits as much as visible diversity
The appointment chain has improved since 2006. Calls are public. Self-nomination is possible. Terms and rotation expectations are stated. Recent rosters identify stakeholder groups, affiliations and terms. Guidance describes target composition and underrepresented countries. Exceptional 2025 arrangements were disclosed, and the 2026 renewal explained the return path.
The remaining weakness is not that the United Nations has a role. The Forum exists because governments asked the Secretary-General to convene it. Final appointment by the Secretary-General is institutionally coherent. The weakness is the ease with which observers can mistake a balanced appointed slate for direct authorisation by global constituencies.
The honest account is stronger. Stakeholders may nominate and recommend. Officials and staff compare candidates and construct a slate under several balance objectives. The Secretary-General appoints. Members exercise individual judgment and are expected to maintain links with communities. Annual evaluation determines whether most may continue. The public rules are less complete when service ends early or when selection tradeoffs must be reconstructed.
That account identifies where accountability belongs. Nominating organisations should explain whom they consulted. Administrators should explain criteria and tradeoffs. The Secretary-General should own appointments and exceptions. Members should disclose affiliations, consultation and conflicts. The Forum should maintain tenure and performance records. No link can borrow legitimacy from the label on another link.
Multistakeholder composition remains valuable. It can prevent one state bloc, industry or professional community from defining the programme alone. It can place very different evidence in the same discussion. It can make an advisory body more perceptive than a homogeneous committee.
But diversity is not delegation. A person can contribute indispensable knowledge without speaking for everyone who shares a category. A nominator can provide credible support without transferring a constituency's consent. The Secretary-General can make a valid appointment without claiming an election occurred.
The MAG earns institutional legitimacy when it makes those limits visible and compensates for them through openness, reasoned selection, consultation and correction. Its members do not need a fictional mandate from the whole Internet. They need a traceable appointment, a clear term, an honest account of whom they can reach, and a known route by which weak service or serious failure can be addressed. That is enough to support authoritative advice without confusing appointment with representation.
The same standard should apply when the roster looks exemplary. A year with perfect numerical balance can still have an unclear nomination path; a year with an imperfect balance can contain unusually strong consultation and evidence. Publication should let observers evaluate both composition and process rather than use one as a proxy for the other.
Permanence makes this more urgent. The IGF is no longer merely a five-year experiment whose informal practices can be excused as provisional. Annual turnover will accumulate into institutional history, and exceptional choices will become precedents cited by future selectors. A stable ledger and explicit authority map would prevent each cohort from inheriting undocumented understandings about who is normally consulted, which organisations count as focal points and when term exceptions are acceptable.
None of these records should turn appointment into a popularity contest. Strong candidates may come from small communities, unpopular positions or countries without established international networks. Transparent selection means explaining why their knowledge and links answer the Forum's needs, not counting endorsements as votes. The purpose of the chain is to make judgment accountable while preserving room for judgment.
That is the final distinction on which the MAG rests. Appointment can create a valid institutional role. Diversity can improve the quality of the group. Consultation can connect members to people beyond the room. Only a traceable combination of all three can justify the confidence placed in those who shape the world's most visible Internet-governance agenda.