CeSIA is proud to announce that it has joined the Partners Community of the Hiroshima AI Process, the Japan-led framework for safe, secure, and trustworthy AI adopted under the G7.
The Partners Community sits alongside the Friends Group, which is countries only, 66 countries plus the EU, as of March 2026. Launched at the Friends Group's first in-person meeting, in Tokyo in February 2025, the Partners Community brings in companies and international organisations to support that work directly. It's meant to help governments understand and implement the Process, get more AI developers reporting under its transparency framework, and build the kind of multi-stakeholder cooperation a government-only group can't manage alone. As of March 2026, it counted 38 organisations, including the OECD, Google, Microsoft, Anthropic, OpenAI, Mila, SaferAI, and The Future Society, alongside Japanese firms like Fujitsu, NTT, and Rakuten.
On September 25, Charbel-Raphaël Segerie will take part in the Friends Group's 11th online meeting on CeSIA's behalf, where he'll give remarks.
This builds on work we've already done inside the Process: in June, CeSIA contributed to version 2.0 of the Hiroshima Process's reporting framework, the only framework where AI developers publicly report, in a comparable format, on how they manage their own risks.