The most urgent task is to collect a wide range of AI good practices and experiences
Yoichi Iida, Director of International Research and Policy Coordination at the Ministry of Internal Affairs and Communications of Japan (MIC-Japan), reflects on the achievements of the GPAI to date, and discusses expectations for the future.
How can the GPAI use its status as an international multi-stakeholder initiative to play a major role in the development of responsible AI?
The GPAI was launched as an international partnership by countries with shared democratic values to promote the development and use of responsible AI based on the OECD’s AI Principles. The vision of the coalition is to provide global leadership for democratic access to AI across society. In order to accomplish this, one of its central roles is to bottom-up dialogue and activities, facilitated by multi-stakeholder engagement, to achieve genuine levels of diversity and inclusiveness across the global community. Enacting the high-level AI Principles endorsed in 2019 has been an important step forward for the entire global community, and the project-led work of the GPAI is expected to take the leading role in demonstrating best practices, providing guidance and issue case studies for the implementation of these AI Principles. Reflecting the spectrum of stakeholder viewpoints is crucial for the GPAI in fulfilling its duty to develop responsible AI.
What do you think of the outcomes achieved by the GPAI and the working groups after this first year?
Looking back at its initial launch 18 months ago, the GPAI has made significant progress in establishing its multi-stakeholder internal governance structure. In this respect, we greatly appreciate the intensive efforts made by the Lead and Incoming Chairs of Canada and France – and others – in building the decision-making process integral to the GPAI organisational structure. They have facilitated synergies between the high-level concept and overall directions presented by the Council, and the individual WG projects with their specific goals. It is also a wonderful development that the GPAI has welcomed so many new member countries. This was preceded by intensive discussions on the membership requirements and criteria; it was crucial that the membership application assessment process was firmly established. Needless to say, the Working Group (WG) projects have made important strides. We place great value on the achievements made by the WGs in creating specific projects that address priority topics, delivering on outcomes and interim reports on progress, of which include the AI in the Workplace Observation Platform, the AI-Powered Immediate Response to Pandemics project and the IP Primer user guide. We absolutely understand that research and analyses can often take longer than initially estimated, and we look forward to seeing more developments over the coming years.
What do you see as the challenges for 2022?
We are delighted that a commendable number of new Member countries have joined the initiative. However, we also foresee potential challenges around the balance of rapid expansion and robust, yet efficient organisational governance. The GPAI fully agrees on the importance of being more inclusive by encouraging higher participation from the countries of the Global South, but we believe that this goal could be achieved by working jointly with other organisations and initiatives outside the GPAI. Japan would be happy to contribute. There may also be ongoing concerns around the fragility of its financial base. The GPAI is a new initiative in its beginning stages, and it is imperative to raise awareness about its unique positioning and importance.
What issues would your country like to see addressed in future Working Group projects?
The most urgent task is to collect good practices and successful (or unsuccessful) examples that can help governments, businesses, the public and all other stakeholders understand and practically apply the lessons learned. From that perspective, all GPAI projects should be informative and provide guidance in the form of practical goals and real-life methodologies. In this way, we believe that all GPAI projects will be helpful and useful for stakeholders, regardless of the topic.
What do you expect from the Paris Summit?
This is the first annual Summit to be held after a full year of activity by all the WGs and other facilities. We look forward to hearing the WG reports and learning how responsible AI has developed over this period. We have very high expectations for how Canada and France will jointly deliver their messages as the founding countries. It is very important that the GPAI presents an integrated message to its multi-stakeholder community by combining the high-level concepts of the Council with the WG project goals. We would particularly emphasise the need for democratic countries to advocate for the growing importance of achieving a human-focused AI society based on shared values, and by promoting the multi-stakeholder approach.