The Centres’ mission
The Centres of the GPAI Expert Community are international, multi-stakeholder entities that design and deliver applied AI projects for the public good. Their work focuses on making artificial intelligence more responsible, trustworthy and safe.
The three Centres are located in Canada, France and Japan, and benefit from national funding and support. They work together to bring together leading AI experts, policymakers, industry representatives, scientists and civil society from around the world to support the responsible adoption of AI. Their initiatives are grounded in respect for human rights and promote inclusion, diversity, gender equality, innovation, economic growth and environmental and social well-being.
The Centres contribute to the integrated OECD–GPAI partnership in a unique way: they translate policy into practice by delivering concrete, applied initiatives that bridge the gap between principles and implementation. By delivering GPAI Associated Projects, the Centres are complementing the core GPAI mission through their exploratory, use-case-based approach. These projects are designed around real-world use cases, delivering tangible, evidence-based outputs and implemented by the GPAI expert community, including AI practitioners, researchers, students, experts, institutional representatives, economic stakeholders, and civil society actors
Three national centres
The Centres of the GPAI Expert Community is a tripartite collaboration within the Global Partnership on AI (GPAI), now part of an integrated partnership with the OECD, bringing together 44 countries and 240+ experts.
There are three Centres, hosted nationally at CEIMIA (Canada), Inria (France), and NICT (Japan).
Their role in developing and deploying responsible AI initiatives is grounded in ethical principles and good governance, fostering innovation and economic growth to contribute to a more resilient society.
The Centres’ methodology
The Centres structure their work around four strategic activities that support the development and delivery of practical projects through a proven, results-oriented cycle:
The innovation workshop
A co-designed process that brings together multi-stakeholder experts to work directly with governments and organisations on concrete AI challenges, developing practical solutions collaboratively.
The lab
A dedicated space for exploratory and experimental projects that advance research, with a strong focus on responsible and safe AI practices.
The incubator
A platform to scale promising initiatives in partnership with governments and sponsoring organisations, supporting the development, piloting and deployment of both practical and ready-to-use solutions.
The academy
A knowledge-sharing hub that develops training programmes, accessible resources and educational modules to help translate the multi-stakeholder approach into effective, real-world implementation.
GPAI Associated Projects: Highlights from the 2026 work plan 2026
The Centres of the GPAI Expert Community develop practical, real-world projects under the GPAI Associated Project brand. These projects inform GPAI’s policy discussions and support consensus-building among members.
Led by experts, the Centres provide a space to test ideas and collaborate with a wide range of stakeholders. Under the 2026 Work Plan, they are supporting nine GPAI Associated Projects.
- Scaling responsible AI solutions
This project helps AI ecosystems deploy and scale responsible AI solutions. Through a flexible mentorship programme led by leading experts, it supports teams in putting responsible AI frameworks into practice within GPAI and beyond. In 2025, the project launched a call for applications for its third global cohort of AI teams and mentors. Activities will continue in 2026.
- Government Data Sharing Roadmap
As AI advances, government-held data has become essential for building trustworthy and effective AI systems. This project helps public institutions share data responsibly by providing practical tools and frameworks. It supports AI innovation in the public interest while protecting privacy, ensuring interoperability and upholding democratic values. In 2026, the project will offer workshops for GPAI members to help accelerate AI adoption in the public sector.
- Virtuous Innovative Approaches and Data Use Collaboration for AI Training (VIADUCT)
VIADUCT brings together data holders and AI developers to address barriers to data sharing for AI training. The project leverages a multidisciplinary approach across technology, economics and law to foster incentivising, trustworthy and compliant data-sharing modalities. VIADUCT aims to produce reports anchored in concrete use cases, with actionable insights and innovative assets for policy and industry stakeholders.
- AI@Work Labs Network
This initiative connects research labs studying how AI is transforming work and employment. It serves as a platform to share knowledge, test ideas and develop joint research. The network contributes evidence to international discussions, including observatories. A 2025 report maps the current landscape of AI and work and highlights the network’s role. In 2026, the network will expand to new labs and launch collaborative research projects.
- Student communities project: Building responsible AI leadership among youth
Launched in 2021 under the GPAI Future of Work Working Group, this initiative has grown into a cross-Centre collaboration connecting students across three continents. Its main activities — SimuvAction (CEIMIA), the AI Grand Challenge (Inria) and Student-led Field Research (NICT) — promote hands-on learning and international exchange.
Throughout the year, Student Dialogues are held online and alongside GPAI events. These sessions allow students to share research, discuss the societal impact of AI and co-create outputs such as blogs, reports, videos, technical tools and policy briefs.
Building on the Tokyo Innovation Workshop, a platform for exploring practical AI policy and implementation, the Centres are launching two new GPAI Associated Projects in 2026.
- Multilingual and multicultural AI: Empowering vernacular voices
This project aims to develop culturally aware datasets and AI models that support multilingual and multicultural exchange. It brings together communities, cultural experts, governments and researchers to promote inclusive governance and ensure AI systems reflect linguistic and cultural diversity.
- Living labs for impact
This initiative will create a virtual, open-access platform to collect, test and share good practices in AI deployment and governance. The platform will serve as a repository for use cases, technical solutions, training initiatives and governance frameworks.
Designed to be open and scalable, it will support collaboration across sectors such as health, education and energy, across regions from local to global, and across different levels of digital readiness. It will encourage cooperation both within the Global South and between the Global South and North.
Get involved
The Centres are always ready to welcome new contributors to their projects.
GPAI members and experts can participate in GPAI Associated Projects by joining project committees, sharing relevant use cases and piloting the tools and solutions developed in each project.
To contribute your expertise and help turn responsible AI principles into practical results, please send us a message.