The OECD has planned to host the Secretariat of the new Global Partnership on AI (GPAI), a recently launched coalition aims at ensuring that Artificial Intelligence is used responsibly, respecting human rights and democratic values. Arrangements for the OECD’s role as host will be finalized in the coming days.

The GPAI will bring together experts from industry, government, civil society, and academia to conduct research and pilot projects on AI. Its objective, as set out by founding members Australia, Canada, the European Union, France, Germany, India, Italy, Japan, Korea, Mexico, New Zealand, Singapore, Slovenia, the United Kingdom, and the United States, is to bridge the gap between theory and practice on AI policy. An example would be looking at how AI could help societies respond to and recover from the Covid-19 crisis. (See launch statement.)

Basing its Secretariat at the OECD will allow the GPAI to create a strong link between international policy development and technical discourse on AI, taking advantage of the OECD’s expertise on AI policy and its leadership in setting out the first international standard for trustworthy AI – the OECD Principles on Artificial Intelligence. The OECD Principles formed the basis of the G20 Principles on AI endorsed at the Osaka Summit in June 2019.

“AI is a truly transformational technology that could play a catalyzing role in our response to Covid-19 and other global challenges provided it is developed and used with trust, transparency, and accountability,” said OECD Secretary-General Angel Gurría. “The launch of GPAI, an initiative grounded in the OECD AI Principles, marks an important step toward this goal. The OECD is looking forward to building powerful synergies between cutting-edge scientific work envisioned by GPAI and the OECD’s AI policy leadership.”

Born out of the Canadian and French G7 Presidencies in 2018 and 2019, GPAI was officially proposed by France and Canada at the Biarritz Summit in August 2019. G7 Leaders then officially welcomed the OECD’s willingness to support their work to advance AI, in line with its Recommendation on AI. The GPAI will initially be comprised of four working groups focused on responsible AI, data governance, the future of work, and innovation and commercialization.

Under the hosting arrangement being finalized, GPAI’s governance bodies, consisting of a Council and a Steering Committee, would be supported by a Secretariat housed at the OECD. The OECD would also be a Permanent Observer to GPAI’s governing bodies and its experts would participate in the working groups and plenary meetings. Inaugural meetings of these groups are expected in late 2020. The GPAI Secretariat would also liaise with Centres of Expertise in Montréal and Paris.

The complementarity of GPAI activities to OECD work should strengthen the evidence base on which the OECD’s policy analysis is developed, and the OECD’s substantive policy work will equally inform discussions in the GPAI’s bodies and working groups. Hosting the GPAI Secretariat will strengthen the OECD’s potential to disseminate and implement its standards and its policy analysis in areas such as data governance, future of work, and diffusion, and productivity.

The OECD’s AI Principles, adopted in May 2019 and now supported by more than 40 countries, comprise five values-based principles for the responsible deployment of trustworthy AI and five recommendations for policy and international co-operation. They offer a guide for designing and running AI systems in a way that puts people’s best interests first and ensuring that AI system designers and operators are held accountable for their proper functioning.

The OECD also operates an online platform – the OECD AI Policy Observatory or OECD.AI – where all players in the AI sphere can share insights and collaborate on shaping AI-related policy. The platform contains data and information on AI trends and policies in around 60 countries and material from partners in academia and the private sector. The Observatory brings together work from across the OECD on AI-related measurement and policy issues and will provide a robust basis for analysis and further use by the GPAI