Should ICANN Hold More Public Referenda on New gTLD Approvals?

The Internet Corporation for Assigned Names and Numbers (ICANN) has, since its inception, been tasked with the stewardship of one of the internet’s most fundamental components: the Domain Name System (DNS). Among its most controversial responsibilities is the approval and delegation of new generic top-level domains (gTLDs), which extend the landscape of digital identifiers beyond the traditional .com, .net, and .org. ICANN’s 2012 new gTLD expansion program added over a thousand new domains, from brand-specific addresses like .google and .apple to generic terms such as .bank, .hotel, and .music. While the move was heralded by some as a breakthrough in innovation, competition, and choice, it also provoked intense debate over fairness, transparency, and the legitimacy of the decision-making process. This has led to a growing chorus of voices asking whether ICANN should implement more robust forms of public input—specifically, binding or semi-binding public referenda—when approving contentious gTLD applications.

At present, ICANN operates under a multistakeholder model that includes inputs from governments, industry stakeholders, civil society, technical experts, and individual users. Policy development is often carried out by working groups under the Generic Names Supporting Organization (GNSO), with opportunities for public comment periods, community forums, and review mechanisms. However, these processes are highly technical, procedurally dense, and often inaccessible to the average internet user or even smaller stakeholder groups without legal or financial expertise. While the process is nominally open, it tends to be dominated by a small circle of repeat players—corporate applicants, intellectual property lobbies, and a few well-resourced civil society actors. In this context, public referenda could offer a more direct, democratic channel for capturing global sentiment on whether certain gTLDs should be approved, especially when they concern sensitive or globally contested terms.

One of the most illustrative examples of the limitations of the current system was the prolonged dispute over the .amazon gTLD. The domain was applied for by Amazon Inc., which sought exclusive use of the term as part of its broader brand strategy. However, the governments of eight South American countries, represented through the Amazon Cooperation Treaty Organization (ACTO), opposed the application, arguing that the term represented a region of cultural, environmental, and geopolitical significance. Despite extensive consultations and Governmental Advisory Committee (GAC) advice, ICANN ultimately awarded the domain to the company, citing procedural consistency and legal obligations. Critics argued that the process prioritized legal form over democratic legitimacy and failed to meaningfully reflect broader public and geopolitical concerns. A public referendum on the .amazon issue—whether open globally or limited to affected regions—might have brought greater clarity, legitimacy, and consensus to the decision.

Another controversy emerged around domains like .health, .islam, and .gay, each of which touched on complex issues of public interest, cultural identity, or minority rights. The .health domain raised questions about health misinformation, data privacy, and commercial exploitation of public health terms. The .islam and .halal domains drew criticism from Muslim-majority countries and religious scholars who feared misrepresentation and theological overreach. Meanwhile, the .gay application sparked debate over whether the domain should be operated as a commercial entity or stewarded by nonprofit groups representing LGBTQ+ communities. In each case, ICANN faced accusations that its adjudication processes failed to account for the perspectives of the very communities the domains were meant to serve. Public referenda, while not a cure-all, could have served as a tool to amplify those voices and ensure that community sentiment was given binding or semi-binding weight in the final decision.

Implementing public referenda within ICANN’s governance framework would not be without challenges. First, there is the question of logistics. With a global internet user base exceeding five billion people, organizing referenda that are secure, representative, and technically robust would require significant infrastructure, verification protocols, and multilingual engagement strategies. There is also the risk of manipulation, astroturfing, or the capture of referendum outcomes by coordinated interest groups, particularly in contentious or high-stakes applications. Moreover, defining the electorate—whether it includes all internet users, residents of specific countries, or members of affected communities—would be a politically sensitive task.

Nonetheless, precedents for direct democratic input exist in other international governance domains. Institutions like the European Union and certain United Nations agencies have developed mechanisms for large-scale public consultations, expert polls, and advisory referenda on matters of global policy. ICANN could draw inspiration from these models to design referenda that are advisory rather than binding, yet still carry significant weight in the decision-making process. For instance, a referendum on whether a new gTLD application should proceed could be triggered when a threshold number of objections or public interest concerns are raised during the comment period. Alternatively, ICANN could pilot referenda on a limited set of highly sensitive domains, using them as case studies for broader democratic integration.

Beyond the procedural mechanics, the introduction of public referenda touches on a deeper philosophical question about the nature of ICANN’s legitimacy. As a private nonprofit corporation incorporated under California law, ICANN is not a state actor, yet it exercises quasi-governmental authority over a critical piece of global infrastructure. Its decisions have wide-reaching economic, political, and cultural implications. In such a context, greater public involvement could help balance the technocratic and commercial biases that often dominate domain name policy. Referenda would not only enhance accountability but also symbolically reaffirm the internet as a global commons, whose governance should reflect the will of its diverse and distributed user base.

Of course, public referenda are no panacea. They can oversimplify complex issues, produce polarized outcomes, and be ill-suited for technical decisions requiring expert input. But when applied selectively—especially to domains involving public goods, cultural heritage, or contested identities—they could provide a vital corrective to ICANN’s historically insular processes. At the very least, the consideration of referenda signals a willingness to explore new forms of participatory governance, beyond the stakeholder meetings and public comment forms that have too often failed to engage or empower the broader public.

As ICANN prepares for future rounds of gTLD applications, with hundreds of new proposals expected to surface, the question of democratic legitimacy will loom ever larger. Whether ICANN chooses to incorporate public referenda, expand community-led review panels, or create new hybrid mechanisms of participatory oversight, the imperative remains clear: the governance of the internet’s naming infrastructure must evolve to reflect not just technical efficiency or market demand, but the values, voices, and diversity of its global users. Without such evolution, ICANN risks drifting further from its founding promise of open, inclusive, and accountable internet governance.

The Internet Corporation for Assigned Names and Numbers (ICANN) has, since its inception, been tasked with the stewardship of one of the internet’s most fundamental components: the Domain Name System (DNS). Among its most controversial responsibilities is the approval and delegation of new generic top-level domains (gTLDs), which extend the landscape of digital identifiers beyond…