Up-to-date developments and analyses of the UN Global Dialogue on AI Governance
Following the first UN process to bring every member state together on artificial intelligence — what it is, when it meets, and what the stakeholder submissions made ahead of the first session said.
What's been said
Coming soonThe submissions to the Dialogue, summarised — the risks they raise and the measures they propose, with verbatim quotes wherever possible. The analysis is being rebuilt and will appear here soon.
Timeline & structure
From the Global Digital Compact to Geneva, and how the Dialogue's two bodies relate.
Latest news
Selected coverage of the Global Dialogue from across the web.
Explore the submissions
Coming soonSearch and filter the submissions by stakeholder, region and tag — and read each one's risks, measures and answers.
FAQ
Short answers to common questions.
Next milestone: first Global Dialogue session · 6–7 July 2026, Geneva (Palexpo)
Official programmeLatest news
All newsUN scientific panel readies its first report for Geneva
Ahead of the 6–7 July session, The National reports on the 40-member Independent International Scientific Panel — co-chaired by Yoshua Bengio and Maria Ressa — finalising the preliminary findings it will present to governments, as concerns mount over jobs, energy use and the pace of AI.
The NationalAnalysis: a day at the UN, and what AI means for humanity
Panel co-chair Maria Ressa recounts the first briefing of UN member states on the scientific panel's forthcoming report — four negotiating groups and 31 states took the floor — and argues that the body's independence from any company or country is what will let governments trust the findings it hands to the July Dialogue in Geneva.
RapplerUN sets out the four themes for the inaugural Dialogue
The UN's official preview confirms the 6–7 July Geneva programme, built around four themes — AI's opportunities and impacts, bridging AI divides, trustworthy AI systems and human rights — with the Independent Scientific Panel presenting its preliminary report on the opening day.
United NationsFrequently asked questions
Common questions about the Global Dialogue, with links to the UN's own pages.
- What is the UN Global Dialogue on AI Governance?
- A United Nations forum where governments and other stakeholders from around the world discuss how AI should be governed. It is designed so that every UN member state — not only the largest economies — has a voice. It convenes periodically; the first substantive session is in Geneva.
- How is it different from the Independent Scientific Panel on AI?
- They are two bodies created by the same 2025 resolution. The Panel is 40 experts who produce evidence-based assessments of AI's risks and opportunities. The Dialogue is the forum where states and stakeholders discuss governance. The Panel informs; the Dialogue deliberates.
- Where did it come from?
- It traces back to the Global Digital Compact, adopted at the UN's 2024 Summit of the Future, which called for a global AI policy dialogue and a scientific panel. General Assembly Resolution 79/325 (2025) then formally established both bodies.
- Is anything it agrees binding?
- No. The Dialogue is a forum for exchange and consensus-building, not a treaty body. Its influence is expected to come through shared principles, best-practice exchange and the interoperability of national and regional rules — a point critics raise when questioning how much it can achieve.
- Who can take part, and how?
- All UN member states, alongside the private sector, civil society, academia and technical communities. Ahead of the first session, the UN ran an open call for written submissions; participation in sessions is handled through the official UN programme and registration.
- What are the written submissions analysed on this site?
- Before the first session, more than 1,500 organisations sent written inputs to the UN. The analysable submissions are summarised here — the risks they raise and the measures they propose. The summaries are AI-generated and each risk and measure is classified automatically against a published ontology; the submitters' own verbatim quotes and the UN's record are authoritative.
- Who runs this website?
- An independent project by CeSIA (Centre pour la Sécurité de l'IA), a non-profit working on AI safety. It compiles public information and analyses public submissions to make the Dialogue easier to follow. It is not affiliated with the United Nations.
Follow the Global Dialogue
Occasional, no-spam updates as it unfolds — new analysis, session outcomes and the documents that matter.