Every AI Policy Ever · open in the interactive register

Published (non-binding, current)Non-binding · Report (government-commissioned)

UN Global Dialogue on Artificial Intelligence Governance: Co-Chairs' Summary (Geneva, 6-7 July 2026)

Record UN-13

Issued by
Co-Chairs of the Global Dialogue on Artificial Intelligence Governance (Permanent Representatives of El Salvador and Estonia), supported by a Joint Secretariat of ITU, UNESCO and the UN Office for Digital and Emerging Technologies
Jurisdiction
United Nations
Level · region
International · Global
Published / adopted
2026-09-08
Language
English

Summary

Official summary of the inaugural UN Global Dialogue on AI Governance (Geneva, 6-7 July 2026, under resolution 79/325), with delegations from 163 countries. The Co-Chairs record convergence on four points: AI's promise must be weighed against risks and capacity built; divides in compute, data and skills shape who can govern AI; human rights, especially children's protection, apply across the life cycle; and no single actor can govern AI alone. The cluster on safe, secure and trustworthy AI found safety frameworks often lack evidence, called for a shared minimum baseline, and highlighted technical standards, open benchmarks, shared evaluation tools, third-party testing and incident reporting.

Safety topics

international cooperationevaluation & testinghuman oversightaccountability & liabilitychild safetyincident reporting

Status notes

A chair's summary, not a negotiated outcome: the Co-Chairs state that the themes it records do not constitute negotiated outcomes or consensus positions. It covers the 6-7 July 2026 session and was released on 8 September 2026. The Scientific Panel presented its Preliminary Report (UN-08) at the session, and the AI Dialogue Partnerships Hub showcased more than 350 commitments, announcements and initiatives. The second Dialogue is to be held in New York on 3-4 May 2027, back-to-back with the Multi-stakeholder Forum on Science, Technology and Innovation for the Sustainable Development Goals.

Sources

Open the official source https://www.un.org/global-dialogue-ai-governance/sites/default/files/2026-09/ai_dialogue_co-chairs_summary.pdf

Secondary source: https://www.un.org/ai-dialogue

Local copy

Archived. A source changed since the previous snapshot; both versions are kept. Last attempt 2026-10-17 13:38:06 UTC

Official source (third-party copy, not published) archive/UN-13/2026-10-11/official.pdf
Retrieved 2026-10-11 19:18:35 UTC
SHA-256 a45d45d19f4d200710dc8701f425e6cdeb357bcb23fe4346101210ab3cc91b8c
Secondary source (third-party copy, not published) archive/UN-13/2026-10-17/secondary.html
Retrieved 2026-10-17 13:38:19 UTC
SHA-256 97fc7720b5690dc316d78ac86419ae4a25c795b319175baf7d1cecebf0a0b871

Verification

B
Another official source, such as an announcement by the issuer, confirms the title, date and status. checked 2026-10-17 · alternative-official-url-fetched
Stored copies (un.org): the Co-Chairs' Summary PDF (18 pages, stored 11 Oct 2026; the archiver found it unchanged on 17 Oct 2026) and the AI Dialogue home page, the record's secondary source, stored on 11 and 17 Oct 2026. The PDF confirms the title, the session dates and venue, the participation figures, the co-chairs, the Joint Secretariat, the four areas of convergence and the cluster findings; it states no issue date. The 17 Oct 2026 copy of the AI Dialogue page says the summary was 'released 8 September 2026', which gives the date at alt-official level because the page is the record's secondary source; it also confirms the 3-4 May 2027 second session. The sentence on the BRICS AI statement was dropped from the notes because no stored source of this record supports it.
Second check: Independent check by Claude Code (claude-sonnet-5-5), 2026-10-17: 6 field(s) supported.

History

  1. 2026-10-17verification.level: A → B — level computed from the evidence by tools/pipeline.py2026-10-17-recheck-international-1 · Claude Code (claude-opus-5-5)
  2. 2026-10-17notes: A chair's summary, not a negotiated outcome: it records discussions and areas of → A chair's summary, not a negotiated outcome: the Co-Chairs state that the themes — Release date added from the page; the 'not a negotiated outcome' remark now quotes the summary's own disclaimer; the BRICS sentence, which no stored source of this record supports, dropped; the STI Forum named as the page names it.2026-10-17-recheck-international-1 · Claude Code (claude-opus-5-5)
  3. 2026-10-17date_published: 2026-09 → 2026-09-08 — The AI Dialogue page (17 Oct 2026 copy) says the summary was 'released 8 September 2026'; the month-only value '2026-09' rested on the PDF's folder name and creation date.2026-10-17-recheck-international-1 · Claude Code (claude-opus-5-5)

Related records

UN AI Scientific Panel & Global Dialogue resolution United Nations · Published (non-binding, current)

UN AI Scientific Panel Preliminary Report United Nations · Published (non-binding, current)

Global Digital Compact United Nations · Published (non-binding, current)

UNGA resolution on safe, secure and trustworthy AI United Nations · Published (non-binding, current)

Citation

Co-Chairs of the Global Dialogue on Artificial Intelligence Governance (Permanent Representatives of El Salvador and Estonia), supported by a Joint Secretariat of ITU, UNESCO and the UN Office for Digital and Emerging Technologies. UN Global Dialogue on Artificial Intelligence Governance: Co-Chairs' Summary (Geneva, 6-7 July 2026). 2026-09-08. https://www.un.org/global-dialogue-ai-governance/sites/default/files/2026-09/ai_dialogue_co-chairs_summary.pdf. Status: Published (non-binding, current); verified 2026-10-17 (level B). Local copy archive/UN-13/2026-10-11/official.pdf retrieved 2026-10-11 19:18:35 UTC, SHA-256 a45d45d19f4d200710dc8701f425e6cdeb357bcb23fe4346101210ab3cc91b8c. Every AI Policy Ever, version 1.18 (snapshot 2026-10-18), record UN-13.

Improve this record

Know of an amendment, a status change or an error? Report news or an error on GitHub; the form opens with this record filled in.