Equity-oriented uses of generative artificial intelligence (GenAI) remain largely underexplored. To address the persistent gender inequities in girls’ representation and participation in computing, we present a case study of an all-girls initiative incorporating two GenAI tools in a European makerspace. We analysed qualitative, multimodal data, including 15 interviews with girls, parents, and practitioners, four reflective sessions, two observation fieldnotes, and 44 GenAI-generated images. Harding’s (1986) multi-level framework and feminist pedagogy provide critical lenses through which to examine and discuss the practice and practitioners’ reflections. Stakeholders generally expressed positive views of the all-girls workshop format. Practitioners enacted three steps to support girls’ participation and engagement in GenAI workshops, namely representation, relationship building, and a responsive approach. However, several issues and challenges emerged in practice, including an inappropriate workshop topic, limited peer collaboration, the inaccessibility of the selected GenAI tools, and language barriers associated with prompt-writing. These challenges revealed underlying issues related to the marginalisation of girls’ voices, agency, and assets. We argue that gender-equity interventions in computing need to move beyond the superficial “painting in pink” of girlification and instead centre girls’ voices, agency, and assets through changes across symbolic, structural, social, and individual levels. The study proposes a girl-centred, multi-level framework and practical implications for the use of GenAI to promote gender equity in computing, while also calling for further research and practice on equity-oriented applications of GenAI. This paper deepens understanding of what it means to centre girls in computing and contributes new empirical and critical insights into how GenAI may be leveraged to advance gender equity and inform the design of related interventions.
Liu et al. (Fri,) studied this question.