Pathways to AI Governance Coordination: Policy Recommendations for the G7 Summit

Executive summary

Artificial intelligence (AI) is having a profound impact on the way we live and work. While more innovative breakthrough to benefit people and society are highly anticipated, we must also acknowledge the risks.

In 2016, as the chairing country of the G7 Ise-Shima Summit, Japan proposed a draft guideline that would serve as a list of principles for AI R&D at the Kagawa-Takamatsu Ministerial Conference on Information and Communications. The guideline catalyzed international attention on the need for AI governance. Discussions continued in subsequent G7 summits, and international organizations, countries and regions, companies, industry associations and civil society groups developed AI principles, such as the OECD AI principles.

In addition to numerous risk assessment tools, several major countries and regions are now proposing regulations for AI in order to advance AI principles into practice. As the number of countries, regions, and application areas utilizing AI expands, international coordination is required to ensure that these tools and regulatory frameworks are developed in a coherent manner that is also consistent with the context of each country, region, and application domain.

Japan is the G7 summit host in 2023. Just as the draft guidelines proposed by Japan in 2016 led to important international discussions, this summit will play an important role in charting the future of AI governance and increasing international coordination and action. Recent developments such as generative AI have sparked legal and ethical debates and highlighted the need for ongoing attention to AI. As the use of AI grows, the public will be much more interested in the governance mechanisms in place. The central question for the G7 at this time is, how do we promote AI innovation to advance national and planetary sustainability and resilience while reducing risks to people and society.

This paper proposes two foundational AI governance policy recommendations that the G7 should respect, and four actions for international coordination by the G7.

Foundational policy recommendations

  1. Compliance with shared fundamental values: The development and use of AI technologies should comply with the fundamental human rights and democratic values underlying existing AI principles
  2. Respect for the context of the application: In putting AI principles into practice, the institutional, social, and cultural context of the application, including the field and region in which the AI technology will be used, should be respected while ensuring commonality as much as possible

Actions for international coordination

i. Establish and adopt international standards: International standards for risk assessment and management frameworks across the AI lifecycle should be established and adopted where possible, in addition to sharing best practices and case studies to promote responsible and safe AI design, development, use and operation across countries, regions and organizations

ii. Promote human-AI collaboration research, policy and practice: Research, policy and practice on the beneficial characteristics of collaborative work between humans and machines should keep pace with rapid AI technology advances and the expansion of human-AI collaborative work

iii. Increase literacy on AI use and governance: Efforts should be strengthened to educate and train people with the skills needed to develop and use AI responsibly, to discuss and practice AI governance, and to empower citizens through literacy

iv. Support discussion forum: In response to the rapid development of AI technologies, multi-stakeholder forums for knowledge-sharing and coordinated action-planning on AI governance in an agile manner should be supported

The full report can be downloaded below.

Policy Recommendation(PDF)