On May 28-29, 2025, ICEED participated in an event hosted by the Japanese AI Expert Support Center of the Global Partnership on Artificial Intelligence (GPAI): the Tokyo Innovation Workshop. This was the first such event since GPAI entered its new phase, coined “GPAI 2.0,” following its integration with the Organisation for Economic Co-operation and Development (OECD).
The event was aimed at stimulating participant engagement to identify key challenges related to AI development and brainstorm potential solutions. Co-hosted by the two other GPAI expert support centers in Canada and France, the Tokyo workshop provided a platform for stakeholders from government, academia, the private sector, and civil society to come together to develop project proposals that will be considered when projects are selected for GPAI’s 2026 work program.
Event attendees were divided into groups under four topics: interoperability of international AI governance frameworks, multilingual and multicultural AI, AI utilization in the Global South and strengthening the domestic and global AI ecosystem, and open-source AI. Over two days, groups participated in open dialogue, discussing the challenges they had faced in their respective countries, regions, and industries, and outlining project proposals to tackle these challenges.
The projects proposed were as follows:
AI utilization in the Global South and strengthening the domestic and global AI ecosystem: Participants reached consensus around the creation of an ‘AI Living Lab,’ which would facilitate cross-country sharing, exchange, and testing of best practices in AI, including use cases and data. The AI Living Lab would be scalable across sectors, geographies, and contexts and would assist in addressing skills and knowledge gaps, while paving the way for community-led AI initiatives and protecting national sovereignty.
Multilingual and Multicultural AI: Participants proposed a multicultural and multilingual AI consortium to address the lack of data in minority languages, which pre–empts a lack of language learning models catering to languages other than English. The consortium would work to enable greater access to and provision of data in minority languages and would actively develop such data sets.
Interoperability of International AI Governance Frameworks: Participants proposed the development of voluntary standardized technical governance tools to facilitate the use and exchange of data in advancing AI innovation globally. This proposal aimed at tackling the challenges posed by the limitations of data use and access by regional governance frameworks and legal protections. An international sandbox to test the tools was suggested as part of the tools-development process.
Open-Source AI: Participants proposed a governance framework to support the safe and responsible development, deployment, and use of open-source AI models. In developing this governance framework, participants suggested a public consultation to identify cross-sector needs and a gap analysis of existing governance tools.
ICEED was encouraged by the quality and ambition of the project proposals developed during the workshop, and hopes that this spirit of collaboration and innovation continues as GPAI moves forward. We will continue to engage with GPAI as it continues its work post-integration, with a view to furthering responsible AI governance and encouraging the advancement of AI through sound policies.