NETmundial 2014 and the Authority of a One-Time Meeting

Summary

• NETmundial's organizers reported 1,480 onsite and remote entities from 97 countries and received 188 written contributions from 46 countries. Those figures demonstrate unusual reach, not formal delegation by 97 states or by the world's Internet users.
• Brazil, CGI.br, and /1net had authority to convene and structure the meeting within their capacities. No treaty, election, membership charter, or universal delegation gave the meeting legislative jurisdiction.
• The process widened access through open submissions, public comments, stakeholder microphone queues, remote participation, and visible drafting. Final editorial choices were still made by selected committees, chairs, and drafting groups under severe time pressure.
• Rough consensus described support within the meeting; it did not mean unanimity. Russia, Cuba, India, and a civil-society representative recorded objections, while many other people and institutions were absent.
• The Multistakeholder Statement expressly called itself non-binding. Its principles could guide, persuade, frame agendas, and supply evidence. They became binding only when a competent institution adopted them under an independent legal, contractual, membership, or policy authority.
• Later reaffirmation at NETmundial+10 and the United Nations General Assembly's decision in 2025 to take note of the newer guidelines expanded reception without retroactively turning the 2014 meeting into a standing global authority.

The meeting began with a crisis of trust, not a grant of power

NETmundial followed a period in which the legitimacy of Internet governance was under extraordinary pressure. The disclosures of large-scale communications surveillance had damaged trust in governments and technology institutions. Brazil's president, Dilma Rousseff, used her 2013 address to the United Nations General Assembly to demand rights-respecting global Internet governance. The Montevideo Statement issued by leading technical organizations called for stronger globalization of Internet coordination. Long-running disputes over the historical United States role in critical naming functions remained unresolved.

This setting gave a Brazilian initiative political energy. It did not give Brazil constituent authority over the Internet. A crisis can justify convening, create urgency, and increase willingness to experiment. It cannot determine whose consent is legally or institutionally sufficient.

The distinction was visible before the meeting. Brazil could speak for its government through ordinary state authority and for its policy positions through diplomacy. CGI.br brought domestic experience with a mixed governmental and non-governmental structure. International partners brought technical standing, networks, money, and convening capacity. Civil-society organizations and companies brought constituencies, expertise, and interests of their own.

These were substantial resources. They authorized the actors to invite, organize, contribute, and persuade. They did not combine automatically into a global sovereign. NETmundial's novelty came from bringing those different capacities together without claiming that one had absorbed the others.

Brazil authorized the invitation, not the world's assent

The official launch notice described the event as a Brazilian initiative organized through CGI.br and international entities associated with /1net. Four committees would support it: a high-level multi-actor committee for political articulation, an executive multi-actor committee for the agenda and proposals, a logistics committee, and a governmental advisory committee open to states wishing to contribute.

That was a serious institutional design for a conference. It established who would prepare the event and divided political, editorial, logistical, and governmental functions. The committees could act because organizers created them and entities accepted the meeting's procedures by taking part.

Their authority ended at the process boundary. A high-level committee could encourage international attendance; it could not certify that every affected population had authorized the meeting. An executive committee could synthesize proposals; it could not make its editorial judgment binding on an absent registry or legislature. A governmental advisory committee could widen state input; it could not convert voluntary attendance into treaty participation.

The phrase "global meeting" described subject and ambition. It did not identify a global electorate. Brazil's status as host gave the process diplomatic credibility and practical capacity, while CGI.br's composition offered a model of institutional pluralism. Neither supplied foreign jurisdiction.

This narrow account does not diminish the conveners. It states the power they actually exercised and prevents later admirers from assigning them a mandate they never claimed.

The entity count measured reach, not representation

The organizer site reports 1,480 entities, including remote entities, from 97 countries. In Internet-governance terms, that was broad and impressive. It brought more perspectives into one process than many treaty negotiations, corporate consultations, or technical meetings manage.

But "from" is not the same as "representing." A civil-society entity from a country does not necessarily speak for its citizens or every nonprofit organization. A business representative does not carry the consent of competitors, workers, customers, or the entire private sector. A technical expert may possess knowledge without a mandate from all network operators. An academic speaks through expertise and argument, not a national delegation. A government official may represent the state while lacking a domestic mandate on every sentence.

Even formal governmental representation would not solve the entire problem. Ninety-seven countries were not every country, state equality is not population equality, and a delegation's authority to attend a meeting does not necessarily include authority to accept a binding instrument. The event did not request treaty credentials or ratification.

Attendance strengthened the statement's claim to have heard diverse views. It gave supporters evidence that the principles were not written by one government or industry circle. It did not permit the arithmetical move from many entities to authorization by everyone affected.

Stakeholder categories opened doors and concealed differences

NETmundial organized contributions and speaking opportunities around governments, the private sector, civil society, the technical community, and academia. The categories helped prevent governments or corporations from occupying all visible space. They enabled a round-robin method in which unlike actors could speak alongside one another.

They also compressed internal diversity. A large platform and a small connectivity provider are both businesses but have different interests and capacities. A human-rights organization, a development nonprofit, and a community network may all appear as civil society. A registry executive, protocol engineer, security researcher, and network operator can enter through the technical category while exercising different authority outside the meeting. Governments vary in regime, size, resources, and accountability.

The label "multistakeholder" does not say how those differences are governed. As Mark Raymond and Laura DeNardis later argued in their typology of multistakeholderism, the important question is not simply how many actor classes are present but what authority relations procedural rules establish among them.

At NETmundial, categories distributed access to microphones and committee places. They did not turn every category into a membership association with an electoral register, instructions, recall, and an appeal route. Category balance was a procedural design choice, not a proof of delegated representation.

The 188 submissions formed an agenda, not a referendum

Before Sao Paulo, the process received 188 written contributions from 46 countries. The executive committee records classified submissions by subject and by the contributor's declared sector. All contributions were to be made available, and a synthesis would inform the draft.

This was a genuine opening. Individuals and organizations outside the committees could place ideas into the public record. The range of submissions gave editors material broader than a closed diplomatic negotiation would have produced. Publication made it possible to compare proposals with the draft.

Still, submissions were self-selected. Forty-six countries were not the world, and a document count did not measure the number of people behind each contribution or the quality of its reasoning. A government paper, a corporate proposal, a coalition statement, and an individual submission each appeared as one document while reflecting very different constituencies and resources.

No process could convert that archive into a statement without editorial judgment. Someone had to group, prioritize, reconcile, omit, and draft. Describing the result as bottom-up should not erase those decisions. Bottom-up input and selected synthesis can coexist. Legitimacy depends on making the conversion visible and contestable, not pretending that the text assembled itself.

The submissions authorized their authors' proposals. They did not function as ballots giving the most frequent phrase a global mandate.

The draft made the editors' power inspectable

The Executive Multistakeholder Committee prepared a draft after reviewing contributions and consulting the high-level committee. It was released for public comment from 14 to 21 April. The comment record and consultation report preserved proposed changes, rationales, and reactions.

That sequence improved procedural legitimacy in two ways. First, entities could see a provisional outcome before arriving. Second, they could identify where editors had combined or excluded positions and propose alternatives. The draft transformed a private editorial act into a public entity.

The consultation window was short, reflecting the compressed timetable. Access to the text did not ensure that every commenter could follow the dense record or travel to the meeting. Editors still had to decide which proposals represented convergence and which would remain contested. The high-level committee's involvement also meant that political judgment entered before the plenary.

None of this invalidates the draft. It identifies the actual authorizing chain: contributors offered inputs; selected committees produced a proposal under organizer rules; public comments supplied further evidence; meeting entities then deliberated. Each stage had a different constituency and effect.

A claim that "the global community wrote the document" collapses those stages. A more accurate claim is stronger because it is verifiable: an international, multi-actor process exposed much of its editorial work and invited challenge before final adoption.

Equal microphone time was a significant but limited innovation

During the working sessions, stakeholder groups used separate microphone queues and speakers were called in rotation. Interventions were generally limited to two minutes. Remote entities also had channels into the plenary. Government officials therefore did not automatically speak first or at greater length, as they often do in intergovernmental meetings.

This mattered. Procedural equality at the point of oral intervention gave non-state entities visible standing and prevented one category from defining the room's priorities. It demonstrated that a government could answer a civil-society advocate, engineer, academic, or business representative within one discussion rather than in separate forums.

The Association for Progressive Communications process study found appreciation for this arrangement. It also reported that 69 percent of survey respondents thought onsite input time was limited public evidence. Only eight hours of the two-day programme were assigned to working sessions, and time limits tightened near the end.

Speaking equality is not decision equality. A two-minute intervention can introduce language or an objection, but editors decide how it changes text. Entities with professional negotiators, prepared wording, and multilingual support can use short slots more effectively than newcomers. Remote access to plenary does not equal access to every drafting conversation.

NETmundial improved one layer of participation without proving equality at all layers. That is an achievement worth preserving and a limit worth naming.

The drafting room was where participation became selection

After plenary interventions, selected chairs, advisers, and drafting groups had to revise the document quickly. Observers could watch parts of this process, but the announced boundary between observation and intervention was not applied uniformly. Some advisers consulted people in the room, and remote entities lacked equivalent drafting access.

The APC study reported mixed views. More respondents thought the drafting was transparent than not, but only 27 percent believed different stakeholder groups were equally represented in the onsite drafting committees. Interviewees described useful openness alongside uneven informal influence and confusion about the consultation rules.

This is the decisive institutional moment. Participation at scale produces more input than a final text can contain. Drafting authority selects among it. Selection is not a defect; it is governance. The problem arises when the final document is presented as if every entity authored every compromise equally.

The editors faced a nearly impossible assignment: preserve momentum, handle legal and technical language, reflect hours of interventions, and find a text that could survive the closing session. Their judgment may have been reasonable. It was not universally delegated.

The process would have been stronger with clearer drafting rules, equal expert support, visible change logs, and a method for testing disputed formulations. Those lessons concern procedural credibility. They do not retroactively make the outcome binding or void. They show why "open meeting" and "authorized final text" are different claims.

Rough consensus was a local rule, not universal consent

NETmundial sought consensus without a formal vote. Entities and observers described the result as rough consensus or a statement supported by most of the room. The method allowed a usable outcome despite unresolved differences and a two-day timetable.

Rough consensus can be appropriate when the product is advisory. It avoids the false precision of counting heterogeneous categories as identical voters. It encourages accommodation and lets objections shape language without always becoming vetoes. Internet technical communities had long used related practices, though NETmundial's political diversity and compressed drafting were different.

The method also requires discipline. Who declares consensus? What weight does a sustained objection receive? How are remote entities counted? Does silence mean assent, exhaustion, lack of access, or acceptance? NETmundial did not have a permanent membership against which those questions could be answered.

The IETF entity account said the editor-led effort sought a statement supported by most entities and recorded objections from Russia, Cuba, India, and a civil-society representative. Their dissent did not prevent issuance of a non-binding statement. It does prevent a claim of unanimity.

Consensus was therefore valid as the meeting's production convention. It showed substantial support among those engaged under the rules. It did not show that every entity, still less every absentee, consented to be governed by the text.

The first sentence supplied the correct authority label

The final Multistakeholder Statement begins by calling itself the non-binding outcome of a bottom-up, open, participatory process. It says the conference was the first of its kind and "hopefully" contributes to the evolution of the Internet-governance ecosystem.

Those words are unusually honest. The statement did not claim to legislate, amend treaties, bind companies, control technical organizations, or speak on behalf of every user. It identified the source of its credibility—the process—and the intended effect—contribution.

The non-binding label also aligned product with method. A voluntary conference with open, self-selected participation, selected editing, no universal credentials, and rough consensus can responsibly issue recommendations. It cannot responsibly create obligations for people who never accepted its jurisdiction.

Advocates sometimes treat non-binding status as a temporary weakness to be overcome through repetition. That misses the design. The principles could travel precisely because recipients remained free to examine, adapt, adopt, or reject them under their own authority. A regulator could use them as guidance. A company could revise policy. A standards body could compare procedures. A court could consider them as context where law permitted. None had to pretend NETmundial was the source of its power.

The first sentence was the meeting's most important accountability control.

Persuasive authority can be substantial without becoming command

Authority is not binary. A document can exert at least four different kinds of influence. It can shape an agenda by naming problems. It can provide reasons and vocabulary. It can serve as evidence that a diverse group found common ground. It can be adopted by institutions that possess binding power.

NETmundial performed the first three directly. It made meaningful participation, accountability, distributed governance, human rights, security, innovation, and a unified Internet part of a common frame. It demonstrated that governments and non-state actors could negotiate principles in one room. Its compromises gave later actors a starting point they did not need to write from nothing.

The fourth effect required a handoff. If a government incorporated a principle into law, the constitution and legislative process supplied authority. If a corporation amended bylaws or policy, corporate governance and contracts supplied authority. If a standards body changed procedure, its membership or consensus rules supplied authority. If an international organization adopted language, its constituent instrument and member decision supplied authority.

The distinction preserves accountability. A government cannot defend coercive policy merely by saying NETmundial wanted it. It must identify lawful power and review. A private institution cannot call itself globally legitimate merely because its preferred principle appeared in Sao Paulo. It must show how affected people participate and obtain remedy.

Persuasion is strongest when it refuses to disguise itself as command.

Human-rights language did not originate the rights

The statement placed human rights at the center of Internet governance, including freedom of expression, association, privacy, accessibility, and protection against unlawful surveillance. This was politically important after the surveillance disclosures and helped resist a purely commercial or technical account of the Internet.

But the rights did not become binding because NETmundial named them. States already had obligations under international and domestic law. Companies and technical institutions could have responsibilities under law, contract, policy, or human-rights frameworks. NETmundial restated and connected those norms to Internet governance; it did not create a global court or new cause of action.

This distinction matters when the statement's language was incomplete. Entities disputed surveillance wording, intermediary liability, copyright, and network neutrality. If the conference were treated as the source of rights, omitted or weakened language might appear to reduce existing protection. It could not. A two-day non-binding meeting had no authority to subtract from treaties or constitutions.

The statement's value was interpretive and political. It asserted that Internet institutions should not treat technical design and governance as outside rights scrutiny. Later decision makers could use that proposition while applying the law that bound them.

The authority chain therefore runs in the opposite direction from a constitutional myth: existing rights gave normative weight to the NETmundial language; the language did not give legal existence to the rights.

Process principles were instructions without a common supervisor

The ten process principles called for multi-actor participation, openness, consensus where possible, transparency, accountability, inclusion and equity, distribution, collaboration, meaningful participation, low barriers, and agility. They became the most portable part of the outcome because they could be applied across institutions.

Their generality was a strength and a limitation. A legislature, technical body, company, intergovernmental organization, and community network can all ask whether participation is meaningful. They will not answer through the same voting rule, disclosure duty, constituency, or appeal.

NETmundial did not establish a certification body to determine compliance. No institution had to submit its process for approval. An organization could cite the principles selectively—celebrating open comments while withholding reasons, or claiming inclusion while participation had no consequence.

The principles therefore operate as standards of criticism. Their force depends on translation into observable controls: advance notice, accessible documents, disclosed interests, balanced selection, response to comments, reasoned decisions, review, funding for participation, and published outcomes. A recipient institution must decide which controls fit its authority and remain accountable for that translation.

A broad statement can make hypocrisy easier to expose. It cannot by itself correct the hypocrisy. That requires an institution-specific reviewer or remedy that NETmundial did not create.

A unified Internet principle did not appoint a global operator

The statement said the Internet should remain globally coherent, interconnected, stable, unfragmented, scalable, and accessible, based on a common set of unique identifiers and allowing the free flow of information. This supplied a powerful public-interest orientation.

It did not place root-zone management, address allocation, routing, interconnection, or protocol standards under NETmundial. Those functions remained with their existing institutions and operators. The meeting lacked credentials, systems, contracts, staff, and continuity duties needed to perform them.

The distinction between principle and operation is especially important here. Many actors can endorse a single interoperable Internet while disagreeing about who authorizes root changes, allocates resources, regulates networks, or responds to abuse. A principle can constrain proposals by requiring advocates to explain fragmentation risk. It does not select the decision maker.

If an operator later cited NETmundial, the citation could show alignment with a widely supported value. It could not replace the operator's actual mandate. If a government invoked the unified-Internet principle to oppose another state's policy, the dispute still required law, diplomacy, technical evidence, and an appropriate forum.

NETmundial spoke credibly about the desired property of the system. It did not acquire custody of the system by describing it.

The IANA transition was adjacent, not authorized in Sao Paulo

One month before NETmundial, the United States government announced its intention to transition stewardship of key IANA functions to the global multi-actor community, subject to conditions and a process. The announcement altered the political atmosphere. It answered a central criticism of unilateral governmental influence and made cooperation in Sao Paulo easier.

The NETmundial roadmap welcomed the transition discussion and called for an open process. Entities debated it, and institutional accounts later linked the meeting's momentum to the transition. Yet NETmundial did not authorize the transfer.

The United States government had to decide what stewardship role it would relinquish. ICANN and the participating communities had to develop a proposal. Corporate accountability reforms, contracts, technical plans, and implementation had to be completed. Those processes had their own participation, approval, and review questions.

The RIPE NCC's contemporaneous account correctly said the transition remained ongoing and that communities would have further opportunities to participate. That is incompatible with a claim that the meeting had decided it.

NETmundial's influence may have been real: it strengthened the political case for a distributed, inclusive transition. Influence should not be inflated into authority. Doing so would erase the institutions that bore operational risk and the decision makers who were accountable for the result.

Marco Civil demonstrated a separate democratic chain

President Rousseff sanctioned Brazil's Marco Civil da Internet at the opening of NETmundial. The moment was symbolically potent. A host country presented a domestic Internet bill of rights while an international meeting debated global principles.

The law's authority did not come from the conference. It came from Brazil's legislative and constitutional processes, including prior public consultation, congressional enactment, and presidential action. People subject to the statute could invoke domestic institutions and courts in ways the NETmundial statement did not provide.

The contrast is analytically useful. Marco Civil had a defined jurisdiction, lawmaker, subject population, operative provisions, and legal remedies. NETmundial had a global audience, recommendations, and persuasive reach. One could inspire the other without their authority becoming interchangeable.

This also protects non-Brazilian entities. Attending a conference in Sao Paulo did not place them under Brazilian Internet law or imply consent to export Brazil's governance model. They could study the process and adopt elements through their own institutions.

The opening ceremony displayed two legitimate forms of action side by side: domestic law and transnational deliberation. Their proximity should not be used to blur their different authorization chains.

Recorded dissent limited the claim and improved the record

At the closing, Russia, Cuba, and India expressed disagreement, and a civil-society representative recorded concerns. Their reasons varied. Governments questioned sovereignty, transparency, democracy, and treatment of their proposals. Civil-society criticism included the weakness or omission of protections relating to surveillance, proportionality, liability, and network neutrality.

The objections did not all point toward the same alternative. That matters. Government dissent cannot be aggregated automatically with civil-society dissent as a unified rejection. Some critics wanted stronger state authority; others feared concessions to states or companies. The statement was contested from opposite directions.

Allowing objections to be heard was a procedural strength. It prevented the consensus declaration from erasing visible disagreement. The record permits later users to distinguish a broadly supported core from unresolved issues.

The dissent also fixes the maximum authority claim. A text opposed by named entities cannot honestly be described as unanimously authorized even within the room. A non-binding consensus can survive objections because it is a statement of broad convergence, not an obligation imposed on objectors.

For absent actors, the case is simpler still. They did not entity because they were not present, not because they assented. Silence outside the meeting has no consent value.

Absence did not create a duty to obey

No global consultation can include every affected person. That practical truth does not make global principles impossible. It changes the kind of claim their authors can make.

NETmundial could say that an unusually diverse process found reasons worth considering. It could invite institutions to explain departures. It could provide a benchmark for advocates and policymakers. It could not tell an absent community that attendance by someone in the same stakeholder category bound it.

This is especially important for users who were present mostly through organizations, governments, or general category labels. Internet users are not a membership body. They do not share one interest, and most had no mechanism to select, instruct, or recall a representative in Sao Paulo. A principle described as being in the public interest still requires evidence and accountable adoption.

An absentee can later consent or become bound through another valid process. A company may accept a contractual term derived from the principles. A citizen may become subject to legislation after democratic enactment. A member may be governed by an organization's rules. Those obligations arise at the adoption stage.

The absence rule is not hostility to global deliberation. It is what keeps deliberation from becoming unaccountable government by whoever can attend.

The later NETmundial Initiative could not inherit authority by name

After the meeting, CGI.br, ICANN, and the World Economic Forum supported a separate NETmundial Initiative intended to help turn principles into practical cooperation. The shared name suggested continuity and momentum.

Institutionally, the initiative was new. It had different sponsors, governance arrangements, resources, and proposed activities. Entities who supported the April statement did not thereby authorize every later structure using the brand. Civil-society and technical actors raised concerns about elite influence, permanent seats, duplication of the IGF, and the initiative's legitimacy.

This episode proves why a one-time meeting cannot serve as an indefinite source of authority. It had no standing membership able to approve successors, elect leaders, amend the statement, license the name, or review institutional drift. Organizers could build a follow-up platform under their own authority. They could not present it as automatically mandated by everyone associated with the 2014 outcome.

The distinction is not semantic. A principle may remain persuasive even when a follow-up institution is poorly designed. Conversely, a useful follow-up activity can earn support through performance without inheriting constitutional status from the original meeting.

Names carry political capital. They do not transfer consent.

Later institutional use created a chain of reception

United States officials, ICANN leaders, Internet Society, regional registry entities, companies, researchers, and civil-society groups quickly cited NETmundial. Some called it historic or evidence that the multi-actor model worked. Later policy documents repeated its process principles.

These citations demonstrate reception, not universal adoption. Institutional praise is also interested evidence. An organization whose preferred governance model was validated had reason to emphasize the meeting's legitimacy. A critic had reason to emphasize omissions and unequal influence. Neither reaction should be treated as a neutral measurement.

A rigorous reception ledger asks four questions. Which exact principle was used? Did the recipient merely cite it, adopt it as policy, or change a binding rule? What authority allowed the recipient to act? What review existed for affected people?

This method can reveal substantial influence. An institution may redesign public comment, disclose selection rules, fund participation, or create appeal because NETmundial gave reformers common language. The change is real. Responsibility remains with the institution that adopted it.

Reception is cumulative and distributed. It can make a non-binding norm widely influential without producing one moment of global ratification. That pattern fits the Internet's institutional structure better than a fictional claim that Sao Paulo legislated once for all.

NETmundial+10 reaffirmed through another voluntary process

In 2024, a tenth-anniversary meeting returned to Sao Paulo. Its final report reaffirmed the ten process principles, proposed more operational guidelines, and drew on 154 online contributions and more than 700 onsite and online entities.

The new statement again called itself non-binding. That continuity is significant. Organizers did not claim that ten years of citation had transformed the original meeting into a legislature. They used another participatory process to renew and elaborate the persuasive framework.

The 2024 entities could express their own support and provide better guidance on implementation. They could not retroactively speak for everyone absent in 2014 or 2024. Reaffirmation showed durability among engaged constituencies and increased the evidence available to later adopters.

It also exposed the difference between a principle and its implementation. The 2024 process said the principles had not been fully implemented and supplied more detailed procedural advice. If the 2014 statement had possessed binding authority, a decade without full implementation would have required a compliance and remedy structure. Instead, the response was further guidance.

That is appropriate for a persuasive norm. It would be inadequate for a law. The repeated non-binding label keeps the category honest.

The United Nations took note rather than ratified

The 2025 WSIS+20 resolution addressed NETmundial+10 with careful language. Paragraph 94 took note of the guidelines as a contribution to strengthening Internet governance through inclusive participation, balanced representation, and openness.

"Takes note" matters. The General Assembly recorded the guidelines and recognized their relevance. It did not incorporate every sentence into a United Nations mandate, require implementation by member states, or transform NETmundial into a United Nations organ.

This later reception strengthened persuasive authority. An advocate can accurately say that the Assembly recognized the guidelines as a contribution. The advocate cannot say the Assembly ratified the 2014 principles on behalf of all states.

Had member states wanted a stronger effect, they could have endorsed specified principles, requested implementation, created reporting duties, or assigned work to a named body. Their chosen verb preserved institutional boundaries.

The 2025 resolution therefore supplies a model for responsible uptake. It recognizes the value of a non-state-inclusive process without laundering its output into universal intergovernmental consent.

A principle's authority should be shown through an adoption ledger

The safest way to use NETmundial is to attach each claim to an institutional chain:

Claimed effect	Evidence required	Decision maker that remains responsible
Broad support in 2014	Participation, contribution, drafting, and dissent record	NETmundial organizers and entities for the accuracy of the claim
Persuasive principle	Public reasoning, fit with evidence, and treatment of objections	The actor citing the principle
Organizational policy	Valid board, membership, management, or community procedure	The adopting organization
Domestic legal obligation	Constitutionally valid legislation, regulation, or judgment	The competent state institution
International obligation	Treaty, customary law, or valid intergovernmental act	States and institutions under that authority
Technical implementation	Standards, contracts, operator decisions, and continuity controls	The responsible technical or operational body

The ledger prevents prestige from substituting for permission. It also prevents recipients from evading responsibility. A regulator cannot blame NETmundial for its law. A company cannot treat the statement as consent from users. An operator cannot replace technical risk analysis with a conference citation.

Used properly, the principles improve reason-giving. They force institutions to explain whether participation is meaningful, decisions are transparent, and power is reviewable. The institution must still provide the remedy.

A one-time meeting can author an argument, not a permanent constituency

NETmundial had no continuing membership, electorate, dues, amendment procedure, disciplinary system, or court. It did not decide who would speak for stakeholder categories at the next meeting because no next meeting was constitutionally required. The 2024 anniversary was separately convened.

This makes a crucial difference between a conference and an institution. A standing body can develop precedent, replace representatives, revise procedures, receive complaints, and be held accountable for implementation. A one-time meeting disperses after issuing its text. Its entities return to institutions where their actual authority differs.

The 2014 process could author an argument about how Internet governance should work. It could not maintain a constituency capable of supervising that argument's use. Later actors had to adopt, reinterpret, or contest it.

That absence may have protected the statement. A permanent NETmundial body could have competed with the IGF, entrenched committee members, or turned broad principles into institutional self-interest. Dispersal kept the principles available to many institutions rather than owned by one.

The tradeoff was accountability. No standing office could answer for later overclaims or hear a complaint that an institution misused the name. The remedy is not to imagine continuing authority. It is to identify each later adopter and its review process.

Substantive merit could not cure an authorization defect

Many of the statement's propositions are persuasive on their merits. Open standards support interoperability. Rights protections constrain both state and private power. Transparent procedures make capture easier to detect. Distributed governance can match decisions to competent institutions. Meaningful participation can reveal costs that insiders overlook. None of those propositions becomes less sensible because the meeting was non-binding.

Substantive merit and authorization answer different questions. Merit asks whether a principle is justified by evidence, rights, technical experience, and the public interest. Authorization asks who may apply it to whom, through which procedure, with what consequence, and subject to what review. A principle can pass the first test and fail the second if an actor imposes it without power.

This separation is particularly important when principles conflict. Openness can collide with security. Privacy can conflict with demands for attribution or access to evidence. Permissionless innovation can create consumer, competition, or safety costs. A unified network can be invoked against legitimate local regulation. Multi-actor participation can slow emergency decisions or give concentrated private power a privileged seat.

NETmundial did not rank every collision. It offered a common vocabulary within which competent institutions could reason. Those institutions still had to identify applicable law, affected constituencies, evidence, tradeoffs, and appeal. They could not resolve a hard case by counting the number of sectors present in Sao Paulo.

The same point protects the principles from opportunistic use. A government cannot cite security and stability while ignoring the statement's rights commitments. A company cannot cite innovation while excluding affected users. A technical body cannot cite distributed governance to avoid public explanation. Selective invocation may be rhetorically convenient, but the non-binding statement provides no enforcement officer to correct it.

The corrective power must come from elsewhere: courts, legislatures, regulators, contracts, membership review, independent oversight, or open technical procedures. NETmundial can strengthen the argument for those safeguards. It cannot substitute for them. Good principles deserve adoption; they do not authorize their own imposition.

The meeting's limits were part of its legitimacy

NETmundial's strongest defenders need not claim that Sao Paulo represented the whole world. Its real achievement is more defensible. Organizers built a fast, comparatively open process at a moment of geopolitical distrust. Stakeholder queues disrupted diplomatic hierarchy. Public submissions and comments exposed disagreement. Editors produced a readable statement. Recorded objectors were not erased. The outcome candidly limited itself.

Those features gave the principles persuasive weight. The same record reveals constraints: unequal resources, self-selection, short timelines, selected drafting authority, absent populations, and no common remedy. A mature account can hold both.

The meeting would become less legitimate, not more, if its diversity were used to bind absentees. Participation is valuable because it improves knowledge, exposes interests, and tests reasons. It is not a magical transfer of authority from everyone who shares a category with an attendee.

The principle that governance should be accountable applies to multistakeholder meetings too. NETmundial met that principle best when it described its own output as non-binding and contributory.

Its authority is therefore easy to state precisely. Entities authorized a meeting outcome under the meeting's procedures. Later institutions authorized whatever they validly adopted. Everyone else received an argument, not an order.

Persuasion should be measured by response, not mythology

The durable question is not whether NETmundial was the constitution of the Internet. It was not. The useful question is what happened when its principles reached institutions with power.

Did an organization disclose how representatives were selected? Did it answer comments rather than count them? Did developing-country entities gain resources and consequence? Did a regulator connect rights principles to enforceable safeguards? Did a technical institution explain public-policy impacts and preserve operational competence? Did a company create review for affected users? Did later international processes cite the statement accurately and record dissent?

These responses can be observed. They allow comparison between rhetorical adoption and institutional change. They also keep the conference in its proper role: a source of common language and public reasons.

The alternative is constitutional mythology. In that story, diverse attendance becomes global representation, rough consensus becomes universal consent, and repeated citation becomes retroactive authority. The story is attractive because it avoids the hard work of obtaining authorization separately in every institution. It is also incompatible with the statement's first sentence.

NETmundial's principles can remain influential without the myth. Their survival over a decade, renewed discussion in 2024, and careful United Nations recognition in 2025 show that non-binding work can travel far. What it cannot do is relieve adopters of the duty to identify their power, constituency, and remedy.

Sao Paulo proved that a one-time global meeting could produce a persuasive settlement among many unlike actors. It did not prove that presence is jurisdiction. The principles earned the right to be heard. They did not acquire the right to command those who never authorized the meeting to speak for them.