Conducted in partnership with Mila, this project seeks to provide the AI ecosystem, in particular the member States of the GPAI, with a framework including concrete tools and recommendations to ensure diversity and gender equality throughout the AI cycle based on existing guidelines, good local practices from around the world, and international human rights and gender equality standards.

AI is generally designed, developed, monitored, and evaluated without systematic gender and diversity approaches, which can lead to negative consequences and obstacles to the development and adoption of responsible AI; for example, in a) disproportionate harm against women and marginalised groups, b) missed opportunities for AI to be more impactful due to lack of consideration of these groups, and c) lack of inclusion also in discussions on AI ethics, regulation and strategies. There is a lack of practical guidance based on evidence that effectively complies with human rights and demonstrates its impact based on internationally-agreed indicators. This project aims to respond to this gap, contributing to a just, responsible, and inclusive AI throughout its full life cycle, in accordance with human rights principles, the OECD Principles on Artificial Intelligence and the UNESCO Recommendation on the Ethics of AI.

This project will, taking a global outlook, generate systemic tools, practices, and recommendations for advancing gender equality, diversity, and inclusion in AI for different stakeholders in the AI ecosystem. Outputs will be based on evidence and clearly defined criteria that are linked with human rights, including those to empower women, girls and marginalised groups; and those to mainstream diversity and gender equality practices through the AI cycle.