International Network of AI Safety Institutes – Joint Mission Statement (renamed International Network for Advanced AI Measurement, Evaluation and Science)

Summary

Mission statement agreed by the ten initial members of the International Network of AI Safety Institutes: Australia, Canada, the EU, France, Japan, Kenya, the Republic of Korea, Singapore, the United Kingdom and the United States. It sets up the Network as a forum for a common technical understanding of AI safety risks and mitigations that supports interoperable principles and best practices, and names four priority areas: research on the risks and capabilities of advanced AI, common best practices for testing advanced AI systems, which may include joint testing exercises, shared approaches to interpreting tests, and inclusion of countries at all levels of development.

Safety topics

Status notes

Launched at the Network's inaugural convening in San Francisco, co-hosted by the US Departments of Commerce and State, with the United States as inaugural chair. The US fact sheet released on 20 November 2024 says the ten initial members agreed the joint mission statement ahead of the convening; the statement's own text dates the launch meeting 21 November 2024. The statement builds on the Seoul Statement of Intent of 21 May 2024 (INT-05). At a meeting in San Diego alongside NeurIPS in December 2025, hosted by NIST's Center for AI Standards and Innovation (CAISI), through which the US takes part, the Network was renamed the International Network for Advanced AI Measurement, Evaluation and Science, and the United Kingdom took on the role of Network Coordinator (announced 9 December 2025). On 13 February 2026 the Network published consensus areas and open questions on practices for automated evaluations. In July 2026, at a meeting in Seoul alongside ICML, it completed its first best-practice guidance document. At their meeting in Paris on 29 May 2026 the G7 digital ministers agreed to start discussions under France's G7 presidency, including members of the Network and the OECD, on the comparability of AI risk-assessment frameworks (G7-04).

Sources

Open the official source https://www.nist.gov/document/international-network-ai-safety-institutes-mission-statement

Secondary source: https://www.gov.uk/government/news/efforts-to-share-best-practices-on-ai-measurement-and-evaluations-driven-forward-through-the-international-network-for-advanced-ai-measurement-evalua

Local copy

Verification

History

1. 2026-10-18related: [{"slug": "seoul-statement-of-intent-ai-safety-science", "id": "INT-05"}, {"slug → [{"slug": "seoul-statement-of-intent-ai-safety-science", "id": "INT-05"}, {"slug — two-way link added: AUS-07, IN-01, INT-19, SG-03 link(s) to INT-112026-10-18-fix-related-links · Claude Code (claude-opus-5-5)
2. 2026-10-18date_published: 2024-11-20 → 2024-11-21 — 2024-11-20 -> 2024-11-21: the mission statement dates itself ('gathered today in San Francisco on November 21, 2024 to launch the International Network'); 20 November 2024 is the release date of the US fact sheet, an announcement about it. A statement's date is taken from the document itself (extract.md step 1), and the notes already give both days.2026-10-18-refresh-level-rule-1 · Claude Code (claude-opus-5-5)
3. 2026-10-17notes: Launched 20 Nov 2024 in San Francisco with the US as inaugural chair. Renamed 'I → Launched at the Network's inaugural convening in San Francisco, co-hosted by the — rewritten from the stored files: the launch, co-hosts and inaugural chair (NIST fact sheet), the two dates (fact sheet 'Released November 20, 2024', statement text 'November 21, 2024'), the Seoul Statement of Intent (statement), the rename and UK coordinator role (GOV.UK), the San Diego meeting hosted by CAISI and the 13 February 2026 publication (NIST news), the July 2026 best-practice document (AISI post), and the G7 sentence corrected to the declaration's words ('we will initiate discussions under France's G7 presidency ... including members of the International Network ..., and the OECD, with a view to improving mutual understanding and comparability of AI risk assessment frameworks'), not a task for France to work with the Network2026-10-17-recheck-international-2 · Claude Code (claude-opus-5-5)
4. 2026-10-17summary: All 10 founding members agreed a joint mission statement. It sets the Network up → Mission statement agreed by the ten initial members of the International Network — rewritten from the mission statement (extra_urls): the ten members, the forum and its common technical understanding, and the four priority areas (research, testing, guidance, inclusion); 'Its work centres on the science of AI testing and evaluation' dropped as a description of the later Network2026-10-17-recheck-international-2 · Claude Code (claude-opus-5-5)
5. 2026-10-17official_url: https://www.commerce.gov/news/fact-sheets/2024/11/fact-sheet-us-department-comme → https://www.nist.gov/document/international-network-ai-safety-institutes-mission — the commerce.gov fact sheet refuses the archiver on every attempt; the mission statement itself, the instrument's own text, is now stored from nist.gov, where NIST's copy of the same fact sheet links it as the 'joint mission statement'2026-10-17-recheck-international-2 · Claude Code (claude-opus-5-5)

Related records

Seoul Statement of Intent on AI Safety Science AI Seoul Summit (2024) · Published (non-binding, current)

G7 Digital & Tech Ministerial Declaration 2026 (Paris) G7 · Published (non-binding, current)

Center for AI Standards and Innovation (CAISI), now CAISSI United States · Published (non-binding, current)

Australian AISI Australia · Published (non-binding, current)

IndiaAI Safety Institute India · Published (non-binding, current)

Five Country Ministerial Communiqué 2026 (AI and national security) Five Country Ministerial (Australia, Canada, New Zealand, United Kingdom, United States) · Published (non-binding, current)

Singapore AISI (Digital Trust Centre) Singapore · Published (non-binding, current)

Citation

Network members; launched by the US Department of Commerce and US Department of State at the inaugural convening in San Francisco. International Network of AI Safety Institutes – Joint Mission Statement (renamed International Network for Advanced AI Measurement, Evaluation and Science). 2024-11-21. https://www.nist.gov/document/international-network-ai-safety-institutes-mission-statement. Status: Published (non-binding, current); verified 2026-10-18 (level A). Local copy archive/INT-11/2026-10-17/official.pdf retrieved 2026-10-17 16:29:59 UTC, SHA-256 7380c0ff55972c0c025266db9259ab8188d218784a78e52df4b81d438a2743fa. Every AI Policy Ever, version 1.18 (snapshot 2026-10-18), record INT-11.

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.