Summary

  • On 15 September, the IGF published 80 selected workshops under five broad themes. Its 7–9 September programme update had organised 359 evaluated proposals and an 80-workshop planning baseline through four thematic areas.
  • The 8 September readout recorded only 76 provisional selections and four remaining places, with priority consideration proposed for leading Infrastructure submissions. The final five-theme page does not disclose a proposal-by-proposal path from that snapshot.
  • The missing record is a versioned disposition bridge: original evaluation theme, score or rank band where publishable, baseline status, balance adjustment, conditional acceptance or replacement, final MAG decision, final public theme and later changes.

The most consequential number on the final IGF 2026 workshop page is not 80. It is five. The page groups the selected workshops under Economy, Innovation & Sustainability; Governance, Cooperation & Institutions; Infrastructure, Access & Technical Foundations; Rights, Inclusion & Human Development; and Trust, Safety & Information Integrity.

That is a sensible reader-facing structure. It is also different from the four thematic areas used in the selection record: Cooperation & Policy; Innovation, Economy & Sustainability; Rights, Ethics & Society; and Infrastructure, Security & Access. The difference does not prove that anything improper happened. It proves something narrower: the published categories cannot be read backwards as if they were the original evaluation architecture.

The Programme Development Update describes 359 workshop proposals assessed by the MAG. It says each proposal entered one of four thematic areas, individual scores were consolidated into overall and thematic rankings, and an 80-workshop planning baseline combined normalised overall ranking at 60% with ranking within a theme at 40%. Thematic groups then reviewed that starting point for quality, substantive gaps, coherence and representation. The document called the result still in progress.

That earlier mechanism has already been reported by BTW. The new fact is the final publication layer. The 15 September announcement confirms 80 selections and five broad themes. Counting the entries on the linked page gives 17 workshops in Economy, 16 in Governance, 15 in Infrastructure, 15 in Rights and 17 in Trust. Those are counts of a final index. They are not a crosswalk from the four evaluation groups.

Seventy-six was a dated decision state

The 8 September readout is unusually useful because it preserves an intermediate state. It records 76 provisionally selected workshops at the end of the day: 28 under Cooperation & Policy, 10 and possibly more under Infrastructure, 23 under Innovation and 15 under Rights. It says the MAG converged on giving leading Infrastructure proposals priority consideration for the remaining four places within the ceiling of 80, with final confirmation deferred to Day 3.

That record should not be stretched beyond its date. It does not prove that the final four places all went to proposals originally evaluated as Infrastructure. Nor does the final total of 15 workshops under “Infrastructure, Access & Technical Foundations” prove a five-place gain from the provisional group. The label changed, and the public sources do not identify the origin of every listed proposal.

The same caution applies to the new Trust theme. Its 17 published workshops may draw from more than one earlier area. Without a proposal-level mapping, a reader cannot tell whether a session moved because its subject was re-described, because programme balance changed, because a condition was met, or because the final taxonomy was designed only after selection. The fifth theme is not evidence of manipulation. It is evidence that evaluation structure and publication structure are different objects.

Evaluation, balance and presentation require separate records

The official submission and review process names three stages: Secretariat screening, individual MAG evaluation, and MAG discussion and finalisation. It says lower-ranked proposals that show promise may receive further consideration to improve programme balance and may be revised. The programme update adds thematic-group deliberation. The final page adds public classification.

These stages exercise different forms of control. Screening decides whether a proposal enters evaluation. Scoring compares proposals under published criteria. A thematic group can recommend a slate that addresses gaps and representation. The MAG approves the workshop programme. The publication surface decides how the finished programme will be legible to readers.

Public consultation belongs in this chain as input, not as approval. The Second Open Consultations and MAG Meeting agenda asked stakeholders to identify considerations for MAG decisions, then scheduled workshop review and approval inside the MAG meeting. Participation can supply evidence, expertise and objections. It does not silently become the authority that selected an individual workshop.

This distinction also limits what selection means. The IGF’s own description says the forum informs those with policy-making power but does not produce negotiated outcomes. A workshop’s inclusion allocates scarce programme attention. It is not an endorsement of every claim, speaker or proposal position.

Publish a disposition bridge, not a reconstruction

The remedy is not to force the five public themes back into four names. The final taxonomy may be better for an audience. The remedy is to preserve the transition.

For every proposal, a disposition bridge should record the immutable submitted version; eligibility result; original evaluation theme; average score or a public-safe rank band where disclosure rules allow it; status in the 80-workshop baseline; thematic-group recommendation; any replacement, merger or conditional acceptance; the reason code for a change; the MAG’s final decision; the final public theme; and every later revision, withdrawal or schedule move.

Not every deliberative note must be public. Personal information, protected discussion and details that would expose reviewers can remain in a controlled audit record. The public projection can still show the proposal identifier, version, stage, responsible body, disposition and bounded reason. The important rule is that a later state must not overwrite the earlier one.

This is an editorial proposal, not an IGF commitment. It follows a simple governance discipline: a participant is not automatically a principal, a ranking is not a recommendation, a recommendation is not approval, and an approved workshop is not an institutional position. Each change in authority needs its own receipt.

The five-theme list makes the programme usable. A disposition bridge would make the selection history testable.

Sources