An applied research agenda for data governance for AI

In our last post, we mentioned that we would be looking to publish an Applied Research Agenda for Data Governance for AI. This document aims to surface some of the great ideas that our Working Group developed as we explored what our two cross-domain projects for GPAI’s 2021 work plan would be. 

We spoke on the two projects in our last blog – more detail on both of them is available now on GPAI’s website, and we are delighted to have engaged the Alan Turing Institute on “Advancing Data Justice Research and Practice” and the Aapti Institute / Open Data Institute on “Enabling data sharing for social benefit through data trusts”. Thanks as ever are due to Working Group members who formed independent evaluation panels to select both delivery partners with the CEIMIA.

The Applied Research Agenda for Data Governance for AI is the product of the few months of reflection and work by the Global Partnership for AI (GPAI) Data Governance Working Group in Spring 2021, during which we developed these ideas for projects and programmes of work that could advance GPAI’s mission, and could be implemented by GPAI’s members and in partnership with others.

To give shape to these ideas, we created a number of concept notes that frame challenges around data governance for AI, and the applied research activities that could address those challenges to get to the action and kinds of impact that we want to see GPAI achieve.

From an initial longlist of nearly 30 ideas, we developed seven detailed concept notes. We were delighted that other GPAI Working Groups adopted a similar approach for developing and articulating ideas for projects, and worked with the GPAI Steering Committee to prioritise our ideas in line with their steer for practical projects that could demonstrate near-term progress and have the potential to be built upon in future years.

Prioritisation is hard, but we are extremely fortunate to have a very collaborative, team-spirited Working Group. We combined elements from across the concept notes to develop two cross-domain projects, and we are very excited to be taking those forward over the next 18 months.

Inevitably though, the necessity to focus our efforts means we think that there are many great ideas in the concept notes we developed within the Working Group that we are not able to take forward this year.

That is why we want to share them with you. We want to celebrate the insight and imagination of our Working Group experts and would love to hear your feedback on their ideas. We want to know what stimulates your own thinking, what chimes with your own priorities, and what ideas they may trigger of your own.

Your feedback on these ideas as they currently stand will help shape the Working Group’s thinking as it develops a pipeline of projects that GPAI’s members and the wider AI community can get behind.

We hope you enjoy them, and we give deep thanks to the Working Group members who have brought so much creativity, insight and meaning to GPAI’s mission. We also thank Jacques Rajotte, CEIMIA and his team for all the support and especially to Ed Teather for his big help and his contribution to making the publication of the Applied Research Agenda for Data Governance for AI possible.

The Data Governance Working Group’s Applied Research Agenda for Data Governance for AI is available to view here. If you would like to get in touch about the Data Governance Working Group’s future work, then you can reach us via GPAI’s Montreal Centre of Expertise, the CEIMIA, on, or tweet us at @JeniT and @majamajab.

AI Wonk Dog
Sign up for OECD artificial intelligence newsletter

Fostering a digital ecosystem for AIInternational co-operation for trustworthy AIShaping an enabling policy environment for AIData GovernanceGPAI

Disclaimer: The opinions expressed and arguments employed herein are solely those of the authors and do not necessarily reflect the official views of the OECD or its member countries. The Organisation cannot be held responsible for possible violations of copyright resulting from the posting of any written material on this website/blog.

Sign up for OECD artificial intelligence newsletter