Dialogues

Participants discussed the upcoming G7 Hiroshima AI Process reporting framework, including initial reactions and questions for the future. The dialogue also included a discussion of a recent UNESCO report on network of national AI supervision authorities. 

The dialogue discussed national AI strategies across Latin America, including in Argentina, Brazil, Colombia, and Mexico and the relation with regional strategies and organizations. 

This dialogue covered the network of international AI Safety Institutes (AISIs), including in the U.S., Canada, EU, and France ahead of the first AISI network meeting in San Francisco. 

Participants discussed the OECD-GPAI integrated partnership, the G7 Hiroshima AI Process, and UN developments, including the Summit for the Future and the Global Digital Compact. 

FCAI leaders presented on a “recipe for global governance,” mapping the existing field of global AI governance institutions, and weighing the strengths of a networked iterative approach that works with and builds on existing international fora or a new international body to coordinate governance work. This discussion was followed by a conversation on the AI Safety Institutes.

This dialogue discussed national AI strategies across Africa, including in Nigeria, Senegal, Egypt, Rwanda, Kenya, and Tanzania, as well as the African Union’s AI strategy. Participants discussed global inclusion in international processes and how FCAI participants can include greater Global South participation in their development of AI policy and foster the UN Sustainable Development Goals. 

Participants ranked areas where international cooperation is needed and discussed which initiatives, if any, they saw as the most promising avenues for progress on international governance. Following this discussion, there was a presentation on the G7 Hiroshima AI Process. 

The dialogue covered standards development and cooperation on AI standards in the Indo-Pacific, transatlantic cooperation on international standards, multinational perspectives on standard-setting, and a presentation and discussion on Brazil’s draft AI bill. 

This dialogue discussed trends in generative AI development and deployment. Then, building on those trends, the participants discussed the implications for public policy and international governance, particularly in thinking through legislation that is technology neutral or AI specific. 

This dialogue gave an overview of the UN Global Digital Compact. A common thread in feedback provided by FCAI participants was the challenge of inclusion and equity, particularly with respect to the Global South. The dialogue also discussed recent CEIMIA research on an AI regulatory policy comparative framework. 

The dialogue focused on AI value chains and general purpose AI. The discussion then turned to a review of the year, including discussion on developments in international AI cooperation, such as the G7, GPAI, and TTC. Lastly, the group discussed the FCAI interim report’s recommendations, went over progress, and previewed the next steps. 

Participants discussed the U.S. AI Bill of Rights, the EU AI Act and conformity assessments, Canada’s AI legislation, and research on the AI compliance landscape. 

Participants in the dialogue discussed the AI Act in the context of the new EU digital agenda, as well as how the EU AI Act might define and scope “artificial intelligence.” Participants also discussed the risk-based approach used in the AI Act. 

The dialogue provided an overview of the U.S. NIST AI Risk Management Framework (RMF), as well as how the RMF might contribute to a broader international framework for risk. 

Participants discussed efforts to develop commitments supporting AI in trade agreements; export restrictions, investment screening, and AI; and what is next for AI in trade agreements. 

Recommendation 15 of our October 2021 progress report called for development of “common criteria and governance arrangements for international large-scale R&D projects.” This roundtable discussed historical international large-scale R&D projects, and then discussed the potential for AI use in future projects, including in privacy-enhancing technologies, climate change, economic measurement, and healthcare. 

Participants discussed the risks and opportunities of engagement with China on AI, including an overview of China’s economic and political trajectory, China’s progress on AI and its implications, and next steps for FCAI participants. 

The dialogue overviewed the FCAI Progress Report, Strengthening International Cooperation on AI, and then discussed international cooperation through domestic AI policy, using Canada as a case study. 

Participants discussed the rationale for international cooperation on AI, and then worked to develop recommendations for international cooperation on AI, including AI regulation and conformity assessments, AI standards, and AI projects. 

Participants compared U.S. and EU AI risk assessment frameworks and then discussed the larger challenge of assessing AI-related risks. 

Participants discussed recent developments on AI regulations and standards, the economic and strategic importance of AI standards, and next steps for standard-setting. 

This dialogue covered the strategic significance of AI, the EU Commission and Parliaments’ approach to data governance and AI, Japan’s approach to data governance and AI, and a discussion on health data and AI. 

Participants discussed the frontiers of AI research and international cooperation on AI research and development, particularly on AI applications. 

The dialogue included a presentation on IEEE standards, and an overview of global standards work, including the CEN-CENELEC AI Focus Group, and the IEC SEG 10. 

Participants discussed transatlantic alignment on assessing AI risk, as well as expanding transatlantic research and innovation around AI. 

The chairman of Brookings and chairman of the Board of CEPS provided introductory remarks, followed by comments from FCAI participants on the need for international AI cooperation.