Summary

  • The IGF’s 2026 redesign combines two intersessional formats under four Policy Networks, while keeping a best-practice component.
  • More than 60 community proposals informed the selection, but the public test is now the trace from participation to specific outputs and their use.

Four desks, one unfinished hand-off

A policy idea can enter a forum as a workshop proposal, a mailing-list comment or a case study from someone implementing a rule. The hard part comes later: who carries it into a draft, what evidence survives editing, and where a decision-maker can use it. In 2026, the Internet Governance Forum has reorganized that middle layer. Best Practice Forums, previously a standalone stream, are being integrated into a single Policy Network framework with a best-practice component.

The Multistakeholder Advisory Group approved four networks: Artificial Intelligence, Cybersecurity, Digital Inclusion and Open Digital Infrastructure. The IGF says the choice followed a review of more than 60 community proposals, many of which raised similar themes. Its stated aim is to streamline the work, improve coordination and strengthen the impact of outputs. Those are aims, not results already measured.

That distinction matters because the IGF is a discussion forum. Its mandate is to discuss digital public-policy issues; the forum says it does not produce negotiated outcomes. A network report can inform policy debate, but it cannot itself impose a rule on a state, an operator or a company. The institutional question is therefore narrower and more practical: can a participant follow an idea from an open group into a clearly attributed, usable piece of evidence?

What changed—and what did not

The old Best Practice Forum format gave people a place to compare experience and identify practices. Policy Networks were already year-round groups built around priority issues. The 2026 change puts the former format inside the latter: a network may combine policy analysis with practical experience-sharing instead of running parallel structures. Dynamic Coalitions remain a separate form of open multistakeholder group; the published description does not say they were merged into the new architecture.

The new model keeps several controls visible. The IGF describes the networks as open and bottom-up. MAG members facilitate; the Secretariat coordinates neutrally and supports final outputs. Participants may shape their own working methods, and each group’s structure can evolve by participant consensus. Mailing lists and regular meetings provide entry points beyond a single annual meeting.

The selection process is a compression point. More than 60 proposals became four network umbrellas. That does not establish that ideas were discarded: the IGF says similar proposals could be adapted into the work programme and invites submitters to stay involved. But the public record needs to make that adaptation legible. A list of four themes tells a reader where work is housed; it does not show which proposed question was carried forward, by whom, or in what form.

One network offers an early test

The AI network provides a specific example of how the architecture could work. Its 2026 focus is the gap between AI-governance commitments and implementation, especially in places expected to apply frameworks they did not write. It meets every two weeks and has two open working groups: one for outreach and case studies, another for analysis and the report. The network says a draft report is planned for public comment in November, with findings intended for the December IGF in Nairobi.

That is a more observable chain than a general promise of “impact”: case contributors, a report outline, a draft, a comment period and a scheduled presentation. Each step can be checked. Whether the evidence represents varied implementation settings, whether comments change the draft and whether the final output identifies its limits will matter more than the new organizational label.

The cybersecurity network shows a different continuity question. It describes itself as the evolution of the 2025 Best Practice Forum on Cybersecurity, carrying prior work into the 2026 Policy Network on securing access and core Internet resources during conflict and crises. Its page says details of the work plan and consultation process will follow. That is evidence of a line of work being carried forward, not proof that all four networks share one timetable or method.

The evidence trail is the product

The IGF’s permanent mandate raises expectations for continuity across meetings, policy networks and other digital-governance processes. Permanence does not itself grant the forum regulatory authority. The forum’s own language remains consultative: its activities develop analytical views that can inform wider discussions.

For an outside reader, the useful record is not just attendance or the number of meetings. It is the chain: the original question, the people able to contribute, the method used to weigh evidence, the draft and revisions, and the destination where a policy-maker can consider the result. If that chain is visible, the new framework can make cross-year work easier to follow. If it is not, four network names may simply repackage a process whose influence remains difficult to audit.

The December meeting in Nairobi is a milestone, not a verdict. The stronger test comes after the reports are published: can readers see what the networks learned, what they could not establish, and how the work moved into a separate decision process?

Sources