The paper examines gender gaps across the AI ecosystem, focusing on women’s underrepresentation in AI research and leadership, gender bias and gaps in training data, and unequal access to and use of AI technologies. It argues that these gaps limit both innovation and economic outcomes, particularly as AI becomes more important for science, technology and economic development. The paper calls for gender-responsive STI and AI policies that support women’s participation not only as users, but also in research, design, decision-making and governance. It also considers the specific challenges faced by women in developing countries and links gender inclusion in AI with wider sustainable development goals.