Summary

  • The Global South is a useful description of structural inequality and collective development interests, but it has no single, stable boundary equivalent to a registry membership roll, national electorate or operator association.
  • The 2005 Tunis Agenda assigned distinct roles to governments, the private sector, civil society, intergovernmental bodies and technical organizations. Its call for full participation from developed and developing countries did not authorize one sector to absorb the others.
  • The Group of 77 and China's 2024 Kampala outcome is a formal state coalition mandate on shared development positions. It is not evidence that every operator, civil-society organization or technical community in its member countries approved each claim.
  • ITU's 2025 data justify urgent development attention: 94% of people in high-income economies used the Internet, compared with 23% in low-income economies, while Africa averaged 36%. Those aggregates reveal inequality but also conceal large differences within countries, between urban and rural users and among operator markets.
  • APNIC's 56-economy region contained more than 10,000 direct members by 2024, with 21% from least developed countries, while Australia, Bangladesh and Hong Kong were the three largest membership locations. No single “Southern operator” profile can represent that mix.
  • A representation audit should identify the claimed constituency, authorization route, sector, geography, operator type, issue, date, sample, conflicts and dissent. Event attendance, fellowship support, regional identity and institutional location are evidence of inclusion, not proof of a universal mandate.
  • NRS can support Southern participation without speaking over it by publishing mandate records, commissioning disaggregated operator evidence, preserving minority statements and renewing representation for each material issue.

A label can be accurate and still exceed its authority

Political language often has to compress. “Global South” can make visible common experiences that a country-by-country list would obscure: colonial extraction, unequal finance, technology dependence, trade constraints, low representation in standards and a digital divide concentrated in poorer economies. The label can form solidarity across regions that do not share a border.

Compression becomes a governance problem when a descriptive category is used as if it were an authorization instrument. The sentence “operators in the Global South need lower participation costs” may be supported by fees, travel budgets and interviews across several regions. The sentence “the Global South has chosen this institution to represent its operators” requires evidence of delegation that the label itself cannot supply.

The distinction applies equally to favorable and unfavorable claims. A speaker cannot infer that Southern operators oppose route-security obligations merely because compliance costs are uneven. Nor can an institution infer that they support a new certification service because the service is presented as development assistance. Operators may value security and entity to governance; support portability and reject a new fee; want regional capacity and distrust the organization offering it.

The discipline is simple: describe conditions with data, describe positions with attributed evidence, and describe representation only with a mandate. A claim may be broadly persuasive without being universally authorized. Saying so does not weaken it. It tells readers what kind of evidence they are hearing.

There is no single official boundary of the South for operator governance

Different institutions group countries for different purposes. The World Bank assigns economies to low, lower-middle, upper-middle and high-income categories using Atlas gross national income per capita, updating the thresholds each fiscal year. The United Nations separately recognizes least developed countries, landlocked developing countries and small island developing States. The Group of 77 is a political coalition with its own membership. Geographic speech often includes Latin America, Africa, much of Asia and parts of Oceania, but usage varies.

These classifications overlap without becoming identical. A high-income Caribbean island can share climate and infrastructure vulnerabilities with lower-income small island states. An upper-middle-income country may contain remote communities with far lower connectivity than the national average. A low-income country may host a sophisticated mobile operator or exchange point. Australia and Japan sit inside APNIC's service region but outside most uses of the Global South. Gulf states complicate income and political categories. China and India carry scale unlike small island economies despite participation in Southern coalitions.

The boundary can also change with the issue. A country may align with developing states on technology transfer and with advanced network markets on security standards. An operator owned by a multinational group may serve low-income customers under local regulation while receiving strategy and capital from elsewhere. A community network in a high-income country may face affordability and remoteness problems similar to networks described as Southern.

A representation claim must therefore name its list. “Operators in 14 surveyed African markets” is auditable. “Members of these three associations” is auditable. “The Global South” without a defined constituency is not.

WSIS protected separate roles rather than creating one Southern voice

The Tunis Agenda remains a useful constitutional reference because it does two things at once. It demands greater inclusion of developing countries and refuses to treat all stakeholders as one undifferentiated actor.

Paragraph 29 states that international management of the Internet should be multilateral, transparent and democratic, with full involvement of governments, the private sector, civil society and international organizations. Paragraph 31 ties legitimacy to full participation of stakeholders from developed and developing countries within their respective roles and responsibilities.

Paragraphs 34 and 35 then distinguish those roles: states hold policy authority for public-policy issues; the private sector has an important technical and economic role; civil society contributes especially at community level; intergovernmental bodies facilitate coordination; international organizations contribute to standards and relevant policy.

That settlement is imperfect and often contested. It nevertheless blocks a common shortcut. A government from a developing country does not acquire the private sector's operational mandate merely because both are Southern. A company does not acquire the state's public authority because it runs critical infrastructure. A technical body does not acquire civil society's rights mandate because its services benefit users.

WSIS participation in Tunis was broad for an intergovernmental summit: the ITU reported delegations from 174 states and entities from hundreds of civil-society and business entities. Attendance provided access and visibility. It did not make one business speaker the representative of all business, or one civil-society intervention the decision of a continent.

The lesson from 2005 is not merely “include the South.” It is “include relevant Southern actors in their actual capacities.”

A government mandate is real and bounded

Governments possess forms of authority that private organizations do not. A minister may negotiate under national law and cabinet instruction. A permanent representative may speak for a state at the United Nations. An intergovernmental resolution can record what participating governments adopted.

That authority is consequential, not fictional. When the Group of 77 and China adopted the Third South Summit outcome in Kampala in January 2024, the text described heads of state and government of member countries acting in solidarity. It called for a stronger Global South position, greater digital inclusion, technology transfer, investment, equal participation of countries and a Global Digital Compact responsive to developing-country priorities. Those are attributable collective state positions.

The mandate remains bounded by actor and subject. It does not show that every national legislature debated each paragraph. It does not show that private operators, independent regulators, civil-society groups or technical communities in all member countries endorsed the same institutional design. Some may support the position strongly; others may disagree; many may never have been consulted.

This boundary should be stated without disrespect. Intergovernmental coalitions need collective language to negotiate structural inequality. Their authority becomes clearer, not weaker, when reports say “the member governments of the Group of 77 and China adopted” rather than “all Southern stakeholders decided.”

The same rule applies to national delegations at WSIS and WSIS+20. A state mandate is evidence for the state's position. Claims about operators require operator evidence.

A regional registry mandate is also real and bounded

Regional Internet registries have contractual members, elected bodies, open technical forums and service responsibilities. They can accurately speak about their own policy, budgets, registry data and member decisions. They may also contribute experience to global debates.

Their service regions should not be mistaken for homogeneous political constituencies. APNIC serves 56 economies across Asia and Oceania. RIPE NCC serves more than 20,000 members across 76 countries in Europe, the Middle East and parts of Central Asia. LACNIC includes Latin America and the Caribbean, regions with different languages, island constraints, market sizes and legal systems. AFRINIC's continental scope contains similarly varied economies and operator structures.

Membership itself is not universal operator coverage. Some networks receive resources through national registries or upstream providers. Some are end users rather than access providers. Some do not vote. Open policy participation can include non-members, while corporate decisions may be reserved to members.

An RIR board elected by participating members has a mandate to govern that corporation under its rules. It does not automatically hold a geopolitical mandate to negotiate for every country in the service region. A policy consensus among meeting entities can be valid for regional resource administration without proving that every network operator affirmatively supported it.

The correct claim names the institutional basis: “members approved,” “the board resolved,” “the open forum reached consensus under its rules,” or “surveyed operators reported.” “The region wants” is rarely precise enough.

The APNIC membership evidence defeats a single Southern operator profile

APNIC's own statistics provide a valuable example of internal diversity. The organization announced its 10,000th direct member in 2024. Its analysis reported that 2,102 members, or 21%, were from least developed countries. It also found that Australia had the largest number of direct members, followed by Bangladesh and Hong Kong. Of the 10,000 members, 6,350 were service providers and 3,650 were end-user organizations.

This is exactly the kind of evidence that should replace rhetorical compression. The same regional institution serves mature high-income markets, very large developing markets, small island economies, least developed countries, national registries, multinational carriers, local ISPs, universities, banks, cloud providers and public bodies.

Their number-resource positions will differ. An established Australian operator may hold a substantial historical estate and focus on security automation. A Bangladeshi access provider may prioritize affordable membership payments, final-pool access, local support and rapid growth. A Pacific island operator may face thin backhaul, disaster exposure and a tiny technical team. An end-user bank may value independent routing but not speak for retail connectivity.

APNIC also notes that address distribution is decoupled from population. In its 2024 membership analysis, China, Japan and South Korea held the largest IPv4 delegations, while membership patterns followed a different order. That means population, addresses, member count and operator need cannot be collapsed into one regional share.

Calling all of these organizations “Global South operators” would erase some members and misdescribe others. The evidence supports targeted coalitions, not one profile.

National registries add representation without universal delegation

Seven economies in APNIC's region have National Internet Registries. These bodies provide local-language services and maintain their own memberships while participating in APNIC under regional rules. They create a legitimate intermediate layer, but the authority chain must still be visible.

APNIC's 2025 annual presentation reported 10,330 direct APNIC members and 15,737 NIR subaccounts. The two populations have different contractual relationships. An NIR can understand local conditions better than a distant regional office and can aggregate technical concerns efficiently. Its staff may have extensive operator contact.

Yet an NIR statement is not automatically a vote by every subaccount holder. The question is how the national institution formed the position: member ballot, board resolution, public consultation, technical committee, staff analysis or executive judgment. Each can be valid for a stated purpose, but they carry different mandates.

The voting structure reinforces the point. APNIC has explained that an NIR receives regional voting weight as one APNIC member according to its tier. Local organizations served by it do not each exercise direct APNIC member votes through that relationship. Aggregation is part of the design.

This does not make the NIR unrepresentative. It means reports should not jump from “the NIR serves these networks” to “every served network authorized this global claim.” Representation must be traced through the national body's own governance.

Connectivity inequality is overwhelming evidence for priority, not unanimity

ITU's 2025 figures establish why development language remains necessary. About 74% of the world's population used the Internet, leaving 2.2 billion offline, most in low- and middle-income economies. Internet use reached 94% in high-income economies but only 23% in low-income economies. The average for Africa was 36%, compared with rates between 88% and 93% in Europe, the Americas and the Commonwealth of Independent States.

These are not marginal differences. They justify prioritizing affordable infrastructure, devices, skills, local content, resilient power and rural coverage. They also support travel funding, remote participation and capacity investment so operators from less connected markets can influence global decisions.

The averages do not prove a shared institutional position. ITU also reports sharp internal divides. Globally, 85% of urban residents were online in 2025 compared with 58% of rural residents. In low-income economies, only 14% of rural residents were online. Africa's urban-rural usage ratio was about 2.6. Men and women, younger and older users, and high- and low-income households experience different connectivity.

An operator serving a capital-city fiber market may face conditions unlike a rural wireless provider in the same country. A mobile incumbent with national spectrum, a community network using unlicensed links and a wholesale fiber company do not infer the same policy from the national penetration rate.

Aggregate inequality establishes who deserves attention. Disaggregated operating evidence establishes what they are asking for.

Income categories are analytical tools, not political instructions

The World Bank's income groups illustrate both the usefulness and danger of classification. For fiscal year 2027, it defines low-income economies at gross national income per capita of $1,175 or less, lower-middle-income economies from $1,176 to $4,635, upper-middle-income economies from $4,636 to $14,375 and high-income economies above $14,375.

These thresholds support comparable analysis and change as incomes change. They do not measure political alignment, network maturity, affordability distribution or institutional trust. A country can cross an income threshold without its rural operators suddenly acquiring high-income economics. Currency changes can move classifications. National averages conceal concentration.

Development categories should therefore be attached to the variable they explain. Income group can help analyze fees as a share of national income. Least-developed status can guide concessionary support. Landlocked status can help explain international transit cost. Small-island status can illuminate route diversity and disaster risk.

“Global South” often combines all of those conditions, but the combination must not become circular reasoning: the institution says it represents the South, defines the South as those needing its representation, then cites the label as proof of consent.

Mandate evidence breaks the circle. A member vote, commissioned delegate, signed association position or documented consultation explains authority independently of the classification.

Operator ownership changes the meaning of regional voice

Operators are not only different in size. Their ownership creates different duties and incentives.

A state-owned incumbent may carry universal-service obligations, national-security duties and government policy direction. A privately owned mobile operator may answer to licence conditions and shareholders. A subsidiary of a multinational group may receive capital and technical standards from a regional headquarters. A cooperative or community network may answer directly to users. A municipal network can be constrained by procurement law. A university network has an academic mission rather than a retail one.

These entities can all operate in the same developing country. A government delegate may understand the public interest but not possess authority to cast the private companies' governance votes. A national operator association may represent major licence holders while excluding community networks and data centres. A civil-society coalition may represent user rights while lacking responsibility for routing operations.

The solution is not to choose one “authentic” Southern voice. It is to preserve the distinction among public authority, operating expertise, commercial interest and community impact. A credible delegation can include all four, with each entity's basis stated.

Claims should also disclose ownership conflicts. A large multinational may advocate a regional position that benefits its subsidiaries. That does not make the position false. Readers should know whether the evidence comes from independent local operators, group affiliates or both.

Number-resource issues divide operators in predictable ways

The idea of one Southern operator interest becomes especially weak when applied to Internet number resources.

An operator with large historical IPv4 holdings may prefer stable recognition and low transaction friction. A post-exhaustion entrant may prioritize affordable transfers, leasing evidence and small-block access. A fast-growing mobile network may rely heavily on address sharing and want operationally practical logging rules. An IPv6-first provider may care more about route security and portable IPv6 than about preserving every IPv4 convention.

A resource-weighted registry fee can appear fair to a small newcomer and expensive to an incumbent with a large estate. A flat fee can appear simple to the incumbent and disproportionate to a rural micro-provider. Strict transfer diligence can protect all members against fraud while delaying a buyer that lacks experienced counsel. Mandatory route authorization can improve collective security while placing a heavier staffing burden on a small network.

Regional identity does not settle these tradeoffs. Neither does income. Two operators in the same low-income economy can occupy opposite sides because one is an incumbent holder and the other is a new entrant.

A representative claim must therefore be issue-specific. An association vote on membership fees does not authorize its delegate on sanctions policy. A consultation on IPv6 training does not prove support for centralized route validation. Mandates should expire or be renewed when the question materially changes.

Speeches are evidence of positions, not of constituency size

Governance debates often rely on speeches because they are public, attributable and timely. They are valuable evidence if interpreted correctly.

At the 2024 Internet Governance Forum opening, a United Nations official urged the forum to bridge the digital divide, serve developing and developed countries, foster cooperation between South and North and strengthen national and regional forums. At an IGF 2025 event focused on the Global South, another official emphasized that billions remained offline and that the divide included devices, skills and safe use, not only physical infrastructure.

These statements establish institutional priorities and a development diagnosis. They do not count operator support for a particular registry model. A keynote can identify urgency without carrying a member ballot. An invitation to cooperate can open space without deciding who speaks for whom.

The same caution applies to panels. A balanced stage can still lack a mandate. A entity may be an excellent expert from a country and not a delegate of its operators. Fellowship funding can correct access inequality, but selection by a conference organizer is not election by a constituency.

Reports should therefore cite speeches as attributed views. They should not convert applause, attendance or geographic biography into authorization. Expertise and representation are both valuable; they are not interchangeable.

The Global Digital Compact renewed inclusion without erasing roles

The Global Digital Compact adopted in 2024 recognizes that digital divides fall heavily on developing countries and calls inclusive participation a cornerstone. It supports an open, global, interoperable, stable and secure Internet and reaffirms that Internet governance remains global and multistakeholder, involving governments, private sector, civil society, international organizations, technical and academic communities and other relevant entities in their respective roles.

That language continues the most defensible part of the WSIS settlement. Developing countries need stronger representation, and non-state stakeholders remain necessary because governments do not operate or experience the whole Internet.

The Compact also calls for more diverse participation from developing countries at the Internet Governance Forum. This is an access commitment, not a transfer of authority to whoever attends. Meaningful participation should improve the evidence available to decisions. It does not create a permanent representative class.

An institution implementing the Compact should ask whether participation reaches different operator types, women, rural networks, small islands, least developed economies and local technical communities. It should publish who contributed and how inputs affected the result. It should not claim that the presence of several Southern entities validates every outcome for the entire South.

Inclusion is strongest when the institution can show both reach and limits.

WSIS+20 demonstrated consultation, not universal assent by every stakeholder

The United Nations General Assembly adopted the WSIS+20 outcome, Resolution 80/173, by consensus in December 2025 after an intergovernmental review that included stakeholder consultations. The preparatory record includes inputs from governments, private sector, civil society, international organizations, technical and academic communities, the Internet Governance Forum and WSIS forums.

This is significant. The outcome has intergovernmental authority and a broader evidence base than a closed diplomatic negotiation. It renewed attention to connectivity, rights, development finance and multistakeholder Internet governance.

The authorization chain remains plural. States adopted the resolution. Stakeholders advised, advocated and contributed through several channels. An informal multistakeholder sounding board helped channel non-governmental perspectives. None of those facts means every stakeholder organization became a party to every paragraph.

That is not a defect. Modern governance frequently combines decision authority with consultative legitimacy. The problem arises only when reports blur them. The accurate description is that governments adopted an outcome informed by stakeholder participation, not that all operators in developing countries issued one instruction.

Future reviews should improve the traceability of input. Which operator submissions supported a proposal? From which countries and business models? Which objections remained? A consultation summary should preserve disagreement instead of converting participation into consensus.

The mandate audit should begin with nine questions

Every claim to represent Southern operators should carry a compact mandate record.

First, who exactly is the constituency: named organizations, association members, licensees in specified markets, registry members, surveyed networks or governments? Second, who authorized the speaker: a member vote, board resolution, statute, ministerial instruction, consultation, survey or personal expertise?

Third, what issue does the authorization cover? Fourth, when was it granted and when does it expire? Fifth, which sectors and operator types are included or absent? A mobile association may not include fixed community networks. A registry may include banks and universities as well as ISPs.

Sixth, what was participation? The denominator, turnout, weighting and decision rule matter. Seventh, what conflicts exist, including employer, funder and parent-company interests? Eighth, what material dissent was recorded? Ninth, what evidence supports the substantive claim beyond the authorization itself?

These questions distinguish four common roles. A delegate carries an explicit instruction. An elected representative acts within a defined office. A consultation entity offers evidence without binding the constituency. An expert speaks from knowledge and should not imply delegation.

The record need not be burdensome. A one-page statement can prevent years of inflated claims.

Sampling is necessary when formal delegation is impossible

Many operator populations have no association capable of issuing a collective mandate. Small providers may be unorganized. Community networks may operate informally. Some companies will not discuss sensitive costs publicly. In those cases, structured research is more honest than invented representation.

A good sample begins with the population: countries, operator types, size bands, ownership, urban or rural focus, fixed or mobile technology, direct or indirect resource access and membership status. Selection should not depend only on conference availability or English fluency.

The report should publish the number invited, number responding, question wording, interview dates and missing groups. Quantitative findings need denominators. Qualitative findings should identify recurring themes without implying statistical prevalence that the sample cannot support.

Minority positions matter. If most large operators support one fee model and small rural networks oppose it, the report should not summarize “regional support.” It should show the distribution and why incidence differs.

Confidential interviews can protect commercial information, but the institution should use an independent researcher and publish the method. Anonymous evidence is still accountable when sampling and verification are clear. An unattributed anecdote selected by an interested institution is weaker.

Research cannot create a binding mandate. It can establish what different operators experience and reduce the temptation to let one visible speaker stand for everyone else.

Participation support should buy access, not allegiance

Travel funding, translation, remote access, childcare, connectivity grants and fellowships can correct structural barriers. They are essential if global meetings otherwise hear only organizations able to fund repeated international travel.

The funding relationship creates a risk. Recipients may be treated as proof that the sponsor represents their region, or may feel pressure to endorse the sponsor's agenda. Even without explicit conditions, selection can favor applicants already aligned with institutional language.

Good safeguards are straightforward. Selection criteria should be public. Reviewers should disclose conflicts. Funding should not require policy agreement. Entities should retain the right to criticize the sponsor. Reports should distinguish sponsored attendance from constituency delegation. Repeat awards should not create a closed circle of professional representatives.

Translation also needs substance. Providing interpretation at a meeting is valuable, but agenda documents, ballots and evidence must arrive early enough for local consultation. A translated speech after the decision does not create participation.

The measure of support is not how many Southern faces appear in a photograph. It is whether new evidence entered the decision, whether entities could disagree safely and whether communities at home could review what was said in their name.

Regional balance on a board is not the same as operator accountability

Boards often use geographic diversity as a legitimacy signal. It is better than a board drawn entirely from one country or market. It remains an incomplete proxy.

A director's nationality does not identify the operators who authorized them. A person from a developing country may work for a global corporation, government, civil-society group or university. Each background brings value but not the same mandate. Two directors from one region can share a country and class while leaving rural, linguistic or small-operator interests absent.

Board reports should therefore separate geographic biography, employer affiliation, selection route and constituency. An elected member director has one basis. An independent skills appointment has another. An advisory seat nominated by an operator association has a third.

No director should be forced to act as the sole voice of a continent. That expectation is both unfair and analytically weak. The institution should gather disaggregated evidence rather than relying on identity to supply every missing perspective.

Geographic balance is a floor for access. Mandate disclosure and operator evidence provide accountability.

Claims should be narrowed before they are amplified

The easiest way to improve public statements is to reduce their scope to the evidence.

Instead of “the Global South rejects high registry fees,” say that a named survey of small providers in specified low- and lower-middle-income markets found fees and international payment costs burdensome. Instead of “African operators support this security model,” say that participating members of named network-operator groups endorsed defined technical controls, while several smaller networks raised staffing concerns.

Instead of “the region selected its representative,” identify the eligible constituency, turnout and voting rule. Instead of “stakeholders agreed at a forum,” distinguish a chair's summary from a negotiated outcome, an audience poll or an adopted resolution.

Narrowing is not timid. It makes evidence portable. Other researchers can test whether the same result appears elsewhere. Operators outside the sample can say whether it fits their conditions. Decision makers can target assistance to the affected group.

Broad rhetoric may attract attention, but precise claims accumulate durable authority. A series of bounded findings can eventually support a strong regional conclusion. A universal claim unsupported by a mandate invites justified distrust.

Dissent is evidence of representation working

Institutions often fear that publishing disagreement will weaken a collective position. The opposite is usually true. A statement that records minority concerns shows that entities were not selected merely to endorse a conclusion.

Dissent can be substantive, procedural or jurisdictional. An operator may oppose the policy. It may support the goal but reject the institution's authority. It may agree nationally but not authorize a global statement. These positions should not be collapsed.

Minority reports are especially important within broad Southern coalitions because material conditions differ. Small islands may prioritize continuity and disaster recovery. Landlocked networks may emphasize transit cost. Large mobile groups may seek scale. Community networks may prioritize local autonomy. A compromise that works for one can burden another.

The record should state whether dissenters were outvoted, whether consensus rules were used and whether objections changed the final text. Confidentiality can protect speakers while still reporting the substance and number of objections.

A mandate that survives visible disagreement is stronger than one inferred from silence.

NRS should treat Southern legitimacy as evidence, not branding

The Number Resource Society can contribute positively if it refuses to market itself as the automatic voice of the Global South. Its standing must be earned issue by issue.

First, NRS should publish a mandate register for every representative statement. The register should name the constituency, decision route, participation, scope, date, conflicts and dissent. Staff expertise should be labeled as analysis, not member instruction.

Second, it should commission disaggregated operator studies. Samples should include direct registry members, national-registry users, upstream-dependent networks, incumbents, new entrants, public operators, community networks, fixed and mobile providers, small islands and landlocked economies. Findings should not be weighted by visibility alone.

Third, NRS should reduce participation cost without purchasing agreement. Translation, remote access and local consultation grants should be governed independently. Entities must be free to reject NRS proposals.

Fourth, institutional claims should be narrow. NRS can say what its members adopted, what surveyed operators reported and what public data show. It should not convert its place of incorporation, leadership biography or development mission into a continental mandate.

Fifth, representation should be renewable. A fee mandate does not become a security mandate. A delegate chosen in 2025 should not speak indefinitely in 2027 without renewed authority.

This approach is less theatrical than broad identity claims. It is also more respectful of the operators whose diversity gives the South its political force.

A 2005-2027 audit should track representative compression

The two-decade study should examine how public claims have compressed constituencies over time. Its source base should include WSIS and WSIS+20 outcomes, Global Digital Compact records, G77 statements, RIR board and election materials, operator association resolutions, speeches, attendance lists and available surveys.

Each statement should be coded for speaker, claimed constituency, authorization, sector, geography, issue, evidence, date and dissent. The study should distinguish formal state outcomes from chair summaries, staff reports, conference declarations, expert remarks and member votes.

The audit should then compare rhetoric with operator and economic data. If a speaker claims a regional consensus on affordability, do member surveys include small and rural networks? If a proposal invokes developing-country need, does its fee design account for income, payment friction and operator scale? If an institution claims broad legitimacy, what percentage of eligible organizations participated?

The objective is not to score every statement as true or false. Some broad language is clearly rhetorical. The objective is to identify where descriptive solidarity becomes an unearned claim of authority.

Historical context matters. The 2005 demand for greater developing-country participation responded to genuine exclusion. The 2024 and 2025 outcomes renewed that demand as digital dependence deepened. The audit should preserve that achievement while asking whether participation channels became more accountable.

By 2027, the result should show which institutions narrowed claims, improved mandate records and incorporated operator difference rather than merely adding the phrase “Global South.”

The 2027 test is a traceable chain from operator to claim

An institution claiming to represent Southern network interests in 2027 should be able to answer direct questions.

Which operators authorized the statement? In which countries and service markets do they work? Are they direct members, national-registry users, customers of members or non-members? What share are small, rural, mobile, fixed, community, public or multinational? What vote, survey or consultation formed the position?

What issue was decided? How long does the authorization last? What participation barriers affected the result? Which material groups were absent? What objections were recorded? Who funded the consultation and the representatives' attendance?

The institution should also connect the position to outcome. Did the proposal lower costs, improve portability, increase route security, preserve service continuity or change participation? Development language should be tested by development incidence, not only by institutional expansion.

No organization will answer every question perfectly. The threshold is not exhaustive unanimity. It is a good-faith, reviewable chain that prevents identity from doing the work of consent.

The strongest Southern mandate may be a coalition narrower than the label: small island operators seeking disaster continuity, post-exhaustion entrants seeking transfer access, African rural providers seeking affordable security support, or NIR members seeking a defined regional reform. Specific coalitions can cooperate globally without pretending to contain everyone.

Three failures should trigger an immediate correction

A mandate record becomes actionable when institutions know which defects require correction rather than another broad statement.

The first is constituency substitution. This occurs when a government position is reported as an operator position, a regional staff view as a member decision, or a conference contribution as continental consent. The correction is attribution: restore the actual speaker and authority, then seek evidence from the omitted constituency before extending the claim.

The second is denominator inflation. An institution may cite all countries in a service region, all names on a membership roll or all attendees at an event even though only a small number participated in the relevant decision. The correction is to publish the eligible population, invitations, responses, organizations represented, weighted votes where applicable and material absences. A small but well-described consultation is more credible than a large implied constituency.

The third is issue migration. A delegate authorized to discuss affordability is later presented as supporting a transfer rule, route-security service or constitutional change. Positions can be related without being interchangeable. The correction is renewed, issue-specific authorization and a clear expiration date.

These failures should affect language before they affect institutional rights. A report can be amended to say “participating operators,” “member governments,” “survey respondents” or “the board” without suppressing the underlying contribution. Repeated inflation is more serious because it indicates that ambiguity is being used strategically. In that case, an independent governance committee should review representation claims and require corrected public records.

The correction standard should also protect contributors. A speaker who was invited as an expert should not be blamed because organizers later described them as a regional representative. Responsibility lies with the institution publishing the inflated claim. Contributors should receive the proposed attribution in advance and be able to state that they spoke personally, professionally or under a defined delegation.

This modest discipline would materially improve Internet governance. It does not demand unanimity, create a veto for every absent organization or prevent political solidarity. It ensures that authority grows from evidence rather than from repetition.

Conclusion: solidarity needs authorization to become representation

The Global South remains necessary language because digital inequality remains severe. ITU's 2025 figures show a 71-point gap in Internet use between high- and low-income economies, billions offline and deep rural and gender divides. Global institutions still need more participation, finance and technical influence from developing countries.

Those facts do not produce one operator position. APNIC's own membership spans 56 economies, more than 10,000 direct members, national registries, mature markets, least developed countries, service providers and end users. The diversity within one region is enough to defeat a universal profile.

WSIS, the Global Digital Compact and WSIS+20 provide a better principle: full participation by governments, private sector, civil society, technical and academic communities in their respective roles. The Group of 77 and China provides another legitimate form: governments acting collectively under an intergovernmental mandate. These forms can reinforce each other without being confused.

The governance failure occurs when regional identity is used to skip the authorization question. A speaker's origin, an institution's location, a conference fellowship or a development mission may add perspective. None proves that all operators delegated their voice.

NRS can be credible by resisting that shortcut. It should publish mandates, sample operating conditions, disclose absences, preserve dissent and renew authority by issue. It should support Southern coalitions while allowing them to remain specific and plural.

The 2005-2027 lesson is therefore not that the Global South lacks common interests. It is that common interests deserve institutions rigorous enough to distinguish solidarity from representation. A label can open the door. Only a mandate can authorize someone to walk through it on behalf of others.

Sources and scope

The article does not claim that every use of “Global South” is misleading or that broad coalitions lack legitimacy. It distinguishes descriptive, political, governmental, institutional and operator mandates. The proposed audit would need multilingual source review, membership and turnout denominators, country- and operator-level sampling, and protection for confidential commercial evidence. Aggregate connectivity and income data establish unequal conditions; they do not establish consent to a specific number-resource institution or policy.