Negotiating Internet Openness

[Note: This essay builds on recent remarks reflecting on fifteen years of work by the Freedom Online Coalition (FOC) to advance a free and open internet.]

Fifteen years into the modern era of Internet governance debates, it is tempting to reach for the language of continuity and to frame the present as a natural extension of a once widely shared vision of a free and open Internet. But that framing no longer holds. The environment in which the Internet freedom agenda emerged has shifted so fundamentally that continuity risks obscuring more than it clarifies. What began as a project grounded in optimism about connectivity, participation, and the expansion of rights, now unfolds in a landscape defined by contestation.

The early Internet was widely understood as a liberation technology. It promised to flatten hierarchies, lower barriers to expression, and connect individuals across borders in ways that would strengthen democratic life. That belief was not naïve and it was grounded in real affordances. The architecture of the network enabled openness while the governance ecosystem, though imperfect, leaned toward interoperability and shared norms. The political moment of the post-Cold War and pre-platform consolidation, allowed for a degree of alignment between technological possibility and normative aspiration.

Today, that alignment has fractured. The Internet is no longer primarily imagined as a shared civic space. It is increasingly treated as an arena of strategic competition, one in which states seek control, firms consolidate power, and technological systems are deployed to advance geopolitical and economic interests. In this context, openness is no longer the default condition but a position that must be argued for, defended, and continually renegotiated.

This shift matters not only at the level of discourse but also at the level of design and governance. When the Internet is framed through the lenses of sovereignty, security, and resilience, the incentives change. Interoperability becomes negotiable, cross-border data flows become conditional and, technical standards become sites of geopolitical influence. What emerges is not a single, coordinated transformation, but a cumulative drift toward fragmentation -- what is often described as the “splInternet,” but which is more accurately understood as a governance project.

The implications are profound. Interoperability is not merely a technical feature; it is a civic condition. It enables communication across jurisdictions, facilitates the circulation of information, and sustains transnational forms of association. As fragmentation deepens, through data localization mandates, sovereign routing ambitions, platform restrictions, and competing standards, the capacity to exercise rights across borders diminishes. The architecture of the network begins to mirror the divisions of the geopolitical landscape.

In this sense, the defense of the open Internet cannot be reduced to a defense of norms alone. It requires attention to infrastructure, to protocol and to the material and institutional conditions that make openness possible. Once borders are hardcoded into the network, they are difficult to undo; and, once interoperability is eroded, the political consequences follow.

At the same time, the terrain of digital governance is expanding beyond the Internet as it was once conceived. Artificial intelligence, platform ecosystems, and digital public infrastructure are becoming foundational layers of contemporary governance. These systems shape access to services, mediate participation in public life, and increasingly structure the exercise of rights. They are not neutral tools; they embed assumptions about authority, accountability, and legitimacy.

Here, the risk is not simply the persistence of existing forms of power, but their transformation. The critique of surveillance capitalism has been well established. But its potential replacement, forms of centralized, state-driven digital control, does not resolve the underlying problem. It simply redistributes it. Without safeguards, the shift from private to public control can result in surveillance statism: systems that are more centralised, less contestable, and equally opaque.

The challenge, then, is to ensure that emerging digital systems are rights-respecting by design. This requires moving beyond ex postaccountability toward ex ante governance. Transparency, privacy, due process, and non-discrimination must be embedded in the architecture of these systems, not as optional features, but as structural constraints. The question is not whether digital infrastructure will shape social and political life but whether it will do so in ways that expand or constrain human agency.

Equally important is the question of legitimacy. The Internet freedom agenda, if it is to remain effective, cannot be perceived as the project of a narrow geopolitical bloc. In a context of increasing polarization, legitimacy is not a secondary concern; it is a strategic imperative. A coalition that does not reflect the priorities and perspectives of the Global Majority will struggle to sustain influence. More importantly, it will struggle to articulate an agenda that resonates across different contexts.

This requires a shift in both representation and substance. Representation must be broadened, not only in terms of participation, but in terms of leadership and agenda-setting. Substance must expand to address issues that extend beyond traditional concerns with censorship and access. Infrastructure gaps, affordability, language inclusion, platform dependency, and data extraction are not peripheral issues; they are central to how digital power is experienced in much of the world. An agenda that does not engage with these realities will remain normatively compelling but politically limited.

Rebuilding the Multistakeholder Model
It is within this broader context that the multistakeholder model must be reconsidered, not as an idealized framework, but as a practical mechanism for governing complexity. The model has long been associated with inclusivity, consensus, and shared stewardship. But these associations, while not entirely misplaced, risk obscuring its more fundamental function.

The multistakeholder model is not a model of harmony. It is a model of structured contestation.

The Internet is governed through a distribution of authority that does not map neatly onto traditional institutional boundaries. Governments legislate and regulate; companies design, deploy, and operate infrastructure; technical communities develop protocols and standards; and, civil society advocates, scrutinizes, and holds actors accountable. These forms of power are distinct, and they are not easily subsumed under a single hierarchy.

Any governance model that privileges one form of authority to the exclusion of others produces distortion. A state-centric model may offer clarity and enforceability, but risks narrowing the space for openness and innovation. A market-driven model may enable scale and efficiency, but often lacks mechanisms for accountability. A purely technical model may optimise for performance, but does not inherently address questions of legitimacy or rights.

The multistakeholder model, at its core, acknowledges this plurality. It does not eliminate power asymmetries, nor does it resolve conflicting interests. What it does, at least in principle, is bring these differences into a shared process by creating a space in which competing claims can be articulated, challenged, and negotiated.

This is its defining feature: not consensus, but negotiation.

Negotiation, in this context, should not be understood as a temporary phase leading to agreement. It is a permanent condition. The interests at stake, sovereignty, profit, technical integrity, rights, are not fully reconcilable. They evolve over time and they intersect in complex ways. The role of the model is not to produce perfect alignment, but to prevent unilateral determination.

In a fragmented geopolitical environment, this function becomes even more critical. As state-led models, promising control, coherence, and strategic autonomy gain traction, the multistakeholder approach can appear slow, diffused, and uncertain. These are not trivial criticisms. They reflect real limitations and a misunderstanding of what is at stake. The alternative to structured negotiation is not efficiency but concentration of power. If the Internet is to remain open, it must remain contestable. Contestability requires mechanisms through which decisions can be influenced, challenged, and revised. It requires the presence of multiple actors with the capacity to intervene. It requires, in other words, a governance model that institutionalises disagreement.

The first challenge is credibility. There is a growing perception that multistakeholder processes are performative in the sense that they provide the appearance of inclusion without altering outcomes. When participation is broad but influence is narrow, the model loses legitimacy. This is not merely a reputational issue as it affects the willingness of actors to engage. If processes are seen as predetermined, participation becomes symbolic.

The second challenge is geopolitical. In a context of rising techno-nationalism, the appeal of state-centric approaches is understandable. They offer clear lines of authority, faster decision-making, and alignment with national priorities. For many governments, particularly those navigating complex security and development pressures, these features are attractive. If the multistakeholder model cannot demonstrate comparable effectiveness, without sacrificing its core principles, it risks marginalisation.

The third challenge is participation. Access to multistakeholder forums remains uneven. Financial constraints, geographic concentration, language barriers, and technical complexity all limit who can engage. This is particularly acute for actors from the Global South, whose perspectives are often underrepresented despite being directly affected by the outcomes. An open process that is structurally inaccessible reproduces the very inequalities it seeks to mitigate. Strengthening the model, therefore, requires more than rhetorical commitment. It requires institutional reform.

First, participation must be made meaningful. This entails moving beyond consultation toward co-decision. Clearer rules of engagement are needed to ensure that contributions from different stakeholders have a tangible impact on outcomes. Mechanisms for accountability, both procedural and substantive, should be strengthened. Frameworks such as the São Paulo Multistakeholder Guidelines point in this direction, emphasizing transparency, inclusivity, and shared responsibility. But principles alone are insufficient; they must be operationalized. Meaningful participation requires capacity. Civil society organizations, particularly in resource-constrained contexts, need sustained support to engage effectively. This includes not only funding, but access to information, training, and networks. Without such support, the model risks privileging those who already have the means to participate.

Second, governance must be decentralized. The concentration of digital governance processes in a small number of global hubs creates structural imbalances. It shapes agendas, influences participation, and reinforces existing power dynamics. Building stronger regional and local ecosystems is essential, not as a substitute for global coordination, but as a complement. Regional institutions can provide context-specific insights, facilitate broader engagement, and serve as conduits between local realities and global norms. In this context, decentralization enhances resilience. A more distributed governance architecture is less vulnerable to capture and better able to adapt to diverse conditions. It allows for experimentation, learning, and the development of contextually appropriate solutions.

Third, the gap between technical design and policy must be bridged. Increasingly, decisions that shape rights are made at the level of code through protocols, standards, and system architectures. Yet the communities that design these systems and those that regulate them often operate in parallel, with limited interaction. This disconnect creates blind spots. Technical decisions may overlook social and political implications, while policy interventions may lack technical feasibility. Bridging this gap requires new forms of expertise and collaboration. Policymakers need a deeper understanding of technical systems while engineers need greater awareness of normative frameworks. Interdisciplinary spaces, where these perspectives can intersect, should be expanded. This is not a peripheral concern; it is central to the future of digital governance.

Ultimately, strengthening the multistakeholder model is not about preserving an institutional arrangement for its own sake. It is about maintaining a mode of governance that can accommodate complexity without collapsing into domination and ensuring that the broader digital ecosystem remains a space where power is subject to constraint.

The open Internet, at this stage, is no longer an inherited condition. It is an ongoing achievement. It depends on choices which are highly technical, political, and institutional as well as on the willingness to accept friction, to engage in negotiation, and to resist the allure of simplified solutions.

There is no return to the early optimism of the Internet age. The conditions that made that optimism possible have changed. But this does not imply inevitability because fragmentation is not preordained and control is not the only outcome. What remains possible is a system in which power is distributed, contested, and accountable. The multistakeholder model, for all its imperfections, remains uniquely suited to this task. Not because it resolves conflict, but because it organises it; and, not because it guarantees openness, but because it keeps openness within reach.

In the end, the choice is not between idealized openness and absolute control but between different ways of structuring power. A system where power is negotiated will always be more demanding. It requires patience, resources, and institutional commitment. But it also offers something that more centralized models do not: the possibility of revision, the space for dissent, and the capacity to adapt. If the Internet is to remain global, interoperable, and rights-respecting, it must remain contestable. And contestability, in a fragmented world, is not a byproduct of governance. It is its purpose.