The Group of Seven (G7) is an intergovernmental forum of seven advanced economies — Canada, France, Germany, Italy, Japan, the United Kingdom, and the United States — together with the European Union as a non-enumerated member. In AI governance it serves as a coordination forum among Western-aligned advanced economies, and is the authoring body of the Hiroshima AI Process.
Type: Intergovernmental forum Members: Canada, France, Germany, Italy, Japan, United Kingdom, United States — plus the European Union as non-enumerated member
Overview
The G7 has no permanent secretariat or legal personality; its presidency rotates annually among members, and each presidency sets the agenda and hosts a Leaders' Summit. Its AI governance work runs primarily through the leaders' and digital/technology ministers' tracks and through outputs delivered under successive presidencies, beginning with Japan's 2023 presidency.
Hiroshima AI Process
The G7's principal AI governance vehicle is the Hiroshima AI Process, launched in May 2023 under Japan's 2023 presidency. Its outputs include:
- The Hiroshima Process International Code of Conduct for Organizations Developing Advanced AI Systems, published 30 October 2023.
- The Hiroshima Process International Guiding Principles, a set of 11 principles that accompany the Code.
- The Hiroshima AI Process Reporting Framework, extended under Italy's 2024 G7 presidency and monitored by the OECD.
On 30 October 2023, G7 leaders issued a statement endorsing the Hiroshima AI Process Comprehensive Policy Framework, which the European Commission describes as resting on four pillars: an analysis of priority risks, challenges, and opportunities of generative AI; the International Guiding Principles for all actors in the AI ecosystem; the International Code of Conduct for organizations developing advanced AI systems; and project-based cooperation in support of responsible AI tools and best practices (Source: https://digital-strategy.ec.europa.eu/en/library/g7-leaders-statement-hiroshima-ai-process). The Code of Conduct and Guiding Principles are voluntary and address organizations developing the most advanced AI systems, including foundation models and generative AI (Source: https://digital-strategy.ec.europa.eu/en/library/g7-leaders-statement-hiroshima-ai-process).
Outputs of the Hiroshima AI Process have been cross-referenced by the EU AI Act, the Singapore Model AI Governance Framework, and the EU GPAI Code of Practice.
OECD Reporting Framework
On 7 February 2025, the OECD launched a reporting framework allowing organizations developing advanced AI systems to report on their efforts to promote safe, secure, and trustworthy AI, monitoring application of the Hiroshima Process International Code of Conduct (Source: https://www.oecd.org/en/about/news/press-releases/2025/02/oecd-launches-global-framework-to-monitor-application-of-g7-hiroshima-ai-code-of-conduct.html). The OECD described it as the first framework allowing companies to provide comparable information on AI risk-management actions such as risk assessment, incident reporting, and information-sharing mechanisms (Source: https://www.oecd.org/en/about/news/press-releases/2025/02/oecd-launches-global-framework-to-monitor-application-of-g7-hiroshima-ai-code-of-conduct.html). OECD Secretary-General Mathias Cormann stated that enabling companies to share their practices would contribute to the responsible development, deployment, and use of advanced AI systems (Source: https://www.oecd.org/en/about/news/press-releases/2025/02/oecd-launches-global-framework-to-monitor-application-of-g7-hiroshima-ai-code-of-conduct.html).
The OECD said developers that had pledged to complete the inaugural framework included Amazon, Anthropic, Fujitsu, Google, KDDI Corporation, Microsoft, NEC Corporation, NTT, OpenAI, Preferred Networks, Rakuten Group, Salesforce, and SoftBank (Source: https://www.oecd.org/en/about/news/press-releases/2025/02/oecd-launches-global-framework-to-monitor-application-of-g7-hiroshima-ai-code-of-conduct.html). Organizations were invited to submit inaugural reports by 15 April 2025, with later submissions accepted on a rolling basis and reports able to be updated annually (Source: https://www.oecd.org/en/about/news/press-releases/2025/02/oecd-launches-global-framework-to-monitor-application-of-g7-hiroshima-ai-code-of-conduct.html). The OECD stated that aligning the reporting framework with multiple risk-management systems was intended to promote interoperability and consistency across international AI governance mechanisms (Source: https://www.oecd.org/en/about/news/press-releases/2025/02/oecd-launches-global-framework-to-monitor-application-of-g7-hiroshima-ai-code-of-conduct.html).
2025 Canadian presidency
Canada held the 2025 G7 presidency and hosted the Leaders' Summit in Kananaskis, Alberta, on 16–17 June 2025, where leaders issued a statement on artificial intelligence for prosperity (Source: https://www.consilium.europa.eu/en/meetings/international-summit/2025/06/16-17/). The G7 Leaders' Statement on AI for Prosperity, dated 17 June 2025, frames a "human-centric approach" to AI and sets out commitments across public-sector adoption, support for small and medium-sized enterprises (SMEs), energy, and partnerships with emerging-market and developing-country partners (Source: https://www.g7.utoronto.ca/summit/2025kananaskis/250617-ai.html).
On public-sector adoption, Canada as presidency launched the G7 GovAI Grand Challenge and announced a series of "Rapid Solution Labs," together with a G7 AI Network (GAIN) to advance the challenge, develop a roadmap to scale AI projects, and create a catalogue of open-source and shareable AI solutions for members (Source: https://www.g7.utoronto.ca/summit/2025kananaskis/250617-ai.html). The statement's annex sets out a G7 AI Adoption Roadmap aimed at SMEs, building on the 2024 Italian presidency's report on driving factors and challenges of AI adoption among companies, and committing to an AI Adoption Blueprint developed in cooperation with the OECD, expanded talent exchanges, and trust-building measures (Source: https://www.g7.utoronto.ca/summit/2025kananaskis/250617-ai.html). On trust-building, the roadmap states the G7 will publish a toolkit for AI deployers and raise awareness of the HAIP Code of Conduct Reporting Framework that the OECD is implementing (Source: https://www.g7.utoronto.ca/summit/2025kananaskis/250617-ai.html).
The statement tasks relevant ministers with delivering a workplan on AI and energy before the end of 2025, citing growing pressure on energy grids from AI adoption alongside potential uses of AI to improve energy efficiency and supply-chain resilience (Source: https://www.g7.utoronto.ca/summit/2025kananaskis/250617-ai.html). It also commits to expanding partnerships with emerging-market and developing-country partners through initiatives including AI for Development, the AI Hub for Sustainable Development, Current AI, FAIR Forward, the Hiroshima AI Process Friends Group, and AI for Public Good, and references the United Nations Global Digital Compact (G7 Leaders' Statement on AI for Prosperity (Kananaskis, June 2025)). The statement reiterates the importance of operationalizing Data Free Flow with Trust (DFFT) through trusted cross-border data flows and references the 2024 G7 Action Plan for a human-centered adoption of safe, secure and trustworthy AI in the world of work (Source: https://www.g7.utoronto.ca/summit/2025kananaskis/250617-ai.html).
The statement's own framing distinguishes it from the Bletchley–Seoul line: it is organized around adoption and prosperity, with trust presented as an adoption enabler — needed "especially for smaller firms deploying powerful technologies—to provide assurance to customers" — rather than as a constraint on development. Its Adoption Roadmap commitments are stated in hortatory rather than binding terms, using "intend," "plan," and "look to," with recommendations governments "can choose to implement" and case studies businesses "can choose to replicate" (G7 Leaders' Statement on AI for Prosperity (Kananaskis, June 2025)).
Relationship to other fora
The G7's AI outputs intersect with several broader bodies. In the G20, whose membership is wider, G7 AI outputs are often referenced in discussions. G7 members overlap substantially with the authorship of the OECD AI Principles, and the OECD monitors the Hiroshima Code Reporting Framework. The UN Global Digital Compact is broader in membership and addressed as complementary to the G7's work. The 2025 Leaders' Statement also references the Global Partnership on AI as a collaborator on AI adoption work (Source: https://www.g7.utoronto.ca/summit/2025kananaskis/250617-ai.html).
Relationships
- related: G7 Hiroshima Code of Conduct for Advanced AI (2023), OECD — Assessing Potential Future AI Risks, Benefits, and Policy Imperatives, The Bletchley Declaration (AI Safety Summit, 1–2 November 2023), Paris AI Action Summit Declaration (2025), EU AI Act (Regulation 2024/1689), International AI Safety Report 2025