International Network for Advanced AI Measurement, Evaluation and Science (NAAIMES) — formerly the International Network of AI Safety Institutes
- Type
- Government body
- Place
- International — Multilateral (10 members)
- Last checked
- 2026-09-22
- Next check due
- 2027-03-21
Ways to reach them
- NIST announcement page (US member, CAISI)
This route is retained as reference only. Its verification is missing, more than 180 days old, or the contact window is not currently open; do not rely on it until it is re-checked.
- UK AISI blog on network outputs
This route is retained as reference only. Its verification is missing, more than 180 days old, or the contact window is not currently open; do not rely on it until it is re-checked.
What it does
Intergovernmental network of state AI evaluation bodies. Renamed in December 2025 from 'International Network of AI Safety Institutes' — the word 'safety' was dropped in favour of 'measurement, evaluation and science'. Verified membership as of 2026: Australia, Canada, European Union, France, Japan, Kenya, Republic of Korea, Singapore, United Kingdom, United States. NOTE: India, Italy, Spain, Netherlands, Ireland, Switzerland and China are NOT members. Publishes joint technical outputs (e.g. Feb 2026 consensus areas on automated evaluations). No treaty basis, no legal personality.
Its powers
convene
Honest assessment
The network has NO standalone website, NO secretariat email, and NO public-facing channel that I could verify. Coordination runs through whichever member hosts a convening; the US CAISI convened the founding meeting (Nov 2024) and the UK AISI has published network outputs. A member of the public cannot address the network — only individual member institutes. Treat 'the Network' as a publication brand, not a contactable body.
Notes
The rename is real and confirmed by NIST and UK AISI primary sources. Any directory still calling it the 'International Network of AI Safety Institutes' is out of date. Do not confuse the US 'CAISI' (Center for AI Standards and Innovation) with the Canadian 'CAISI' (Canadian AI Safety Institute) — the acronym collision is genuine.