IGF 2022 Open Forum #29 GPAI: A Multistakeholder initiative on trustworthy AI

Tuesday, 29th November, 2022 (10:45 UTC) - Tuesday, 29th November, 2022 (11:45 UTC)
Caucus Room 11

Tutorial - Auditorium - 30 Min


Built around a shared commitment to the OECD Recommendation on Artificial Intelligence (AI), the Global Partnership on Artificial Intelligence (GPAI) brings together engaged minds and expertise from science, industry, civil society, governments and international organisations. It aims to bridge the gap between theory and practice on AI by supporting cutting-edge research and applied activities on AI-related priorities. 

Initially conceived by Canada and France during their G7 presidencies, the GPAI initiative now has 25 member countries. As an expertise-based initiative, GPAI undertakes projects on specific AI issues in order to support and guide the responsible development, use and adoption of AI that is human-centric and grounded in human rights, inclusion, diversity and innovation.

During this introductory session, GPAI representatives leading the initiative from government and the Secretariat will highlight how the initiative aims to advance insight on important questions.


Global Partnership on Artificial Intelligence (GPAI)
Government of Japan

  • Yoichi Iida, Government of Japan, Deputy Director General for G7/G20 Relations, Ministry of Internal Affairs and Communications
  • Elizabeth Thomas-Raynaud, Head of the GPAI Secretariat, OECD 
Onsite Moderator

Roberto Zambrana

Online Moderator

Hyebin Hong, GPAI Secretariat, IO Stakeholder Group


Hyebin Hong, GPAI Secretariat, IO Stakeholder Group


17. Partnerships for the Goals

Targets: GPAI’s work is to promote a global partnership to share values, knowledge and best practices of AI grounded in a human-centric manner. The GPAI would like to use the opportunity of Open Forum to expose its work to stakeholders and potential member states who have the same values of fostering trustworthy AI. In light of this, the discussion about GPAI’s work aligns with SDG 17. Addressing the impact of TVEC (terrorists and violent extreme content) on social media is important to prevent any possible violence, terrorism and crime which addresses SDG 16.

Key Takeaways (* deadline 2 hours after session)

- To build a safe, healthy and prosperous future, we need to discuss how humanity can live and work harmoniously with AI. It requires collaboration and coordination of diverse stakeholders, such as governments, industry, civil society and academia to unfold the full potential of AI to serve the good of society.

- GPAI has its values in demonstrating powerful synergies by embracing multidisciplinary perspectives from governments, institutions and experts to promote the efforts on developing and deploying human-centric and responsible AI.

Session Report (* deadline 26 October) - click on the ? symbol for instructions

The session was a 30 minute-dialogue moderated by two speakers.

Elizabeth Thomas-Raynaud, the Head of the GPAI Secretariat introduced GPAI’s mission, structure, Experts Working Groups and the different forms of participation in GPAI. Yoichi Iida, Director of International Research and Policy Coordination, at the Ministry of Internal Affairs and Communications of Japan (MIC-Japan) stressed the GPAI’s growth and diversity in promoting its work toward shaping a responsible AI society.

Summary of talking points

  1. GPAI’s mission is to bring governments and experts together to support and guide the responsible adoption of AI grounded in the shared principles from the OECD AI recommendation. Founded in 2020 with 15 Members, GPAI has almost doubled its membership with 29 Members in 2022. On 22 November in the context of the GPAI summit in Tokyo, GPAI welcomed four new Members: Argentina, Senegal, Serbia and Türkye. GPAI produces project outputs under the four topic areas – Responsible AI, Data Governance, Future of Work and Innovation and Commercialisation. The projects are the responsibility of the GPAI Experts who are from a wide range of sectors and a diversity of countries.
  2. Membership in GPAI is open to countries, including emerging and developing countries that endorse and share the values reflected in the OECD Recommendation on AI and have proactive role in advancing responsible AI. Experts wishing to participate in GPAI can do so through a nomination by GPAI Member countries or by the process of self-nomination. The call for self-nominated Experts opens once a year and experts from all around the world are welcome to apply.
  3. AI has progressed to the extent of bringing innovation to industries, transforming our daily lives and providing technology solutions to the most pressing issues of our time. However, this transformative potential can come with challenges when left unchecked. GPAI, despite its infancy, plays an important role in strengthening the shared values and efforts to shape a broader responsible AI community across the globe. To do this, GPAI gathers leading AI experts to produce impactful and useful AI projects in view of helping Members ensure AI development and deployment that elevates humanity.

Summary of interaction

  • Potential risk of overlapping with other initiatives: GPAI was not intended to duplicate activities taking place in other organizations on policy making. GPAI focuses on leveraging insights and expertise from the experts for the applied AI projects that aim to benefit governments in pursuing their AI discussions and implementations.
  • Criteria for Membership: GPAI is open to developing and emerging countries and its strong importance lies in like-mindedness based on the OECD AI principles.
  • Domestic priorities vs. Global priorities: There’s a strong willingness of the majority of GPAI members in contributing expertise and insights from the national AI institutes and their own experiences. Their domestic priorities and interests are shared for the development of GPAI activities and projects. Similarly, insights from international cooperation are contributing to the domestic work on AI. There are mutual benefits from that exchange.
  • A Voluntary initiative: GPAI is a voluntary organization and there is no treaty requirement. GPAI promotes project-based efforts for the development and deployment of responsible AI, while the OECD focuses on discussions of policymaking and governance framework around the topic. They should both go hand in hand to promote the OECD Recommendations on AI which are non-binding. The discussion of binding regulations is left at the government-level policy forums. GPAI distinguishes its work from binding AI regulation and emphasizes the collaborative efforts to facilitate responsible and human-centred use of AI across the globe.