Hiroshima AI Process — G7 code of conduct and reporting framework for advanced AI

The Hiroshima AI Process (HAIP) is the G7's initiative for international governance of advanced AI systems, launched at the G7 Hiroshima Summit in May 2023 under Japan's presidency. It produced the International Guiding Principles on AI and the International Code of Conduct for Organizations Developing Advanced AI Systems (October 2023) — voluntary commitments on risk assessment, incident reporting, transparency, and information sharing for developers of the most advanced AI systems.

In February 2025 the OECD operationalized the Code of Conduct by launching the HAIP Reporting Framework at the Paris AI Action Summit — the first global framework through which companies report, in comparable form, on their AI risk management practices. A streamlined version 2.0 of the reporting framework followed on 28 May 2026, launched at a Tech7 event alongside the G7 Digital and Tech Ministerial Meeting in Paris, designed to make participation practical for small and medium-sized enterprises; more than 50 organizations pledged to submit reports under the new framework, with submissions requested by 1 September 2026 for the next analytical review.

Canada participates as a G7 member, and its 2025 G7 presidency carried the work forward: the leaders' statement on AI for Prosperity issued at the Kananaskis Summit commits to building on the Hiroshima AI Process to strengthen trust in AI through voluntary, standards-based frameworks.