Artificial intelligence is often presented as a technological breakthrough driven by algorithms, large language models, and computing infrastructure. Far less attention is given to the human labour and everyday data practices that make AI systems possible and determine who benefits from them.

Drawing on research with domestic workers in Nairobi Kenya, this session explores how invisible forms of digital labour shape AI infrastructure. Domestic workers increasingly interact with AI mediated recruitment platforms, messaging applications, digital identity systems, payment technologies, and algorithmic management tools. These interactions generate data that are rarely recognised as labour despite becoming part of the digital ecosystems that support AI development and deployment.

The session introduces a feminist lens for understanding AI infrastructure to explore questions of consent, data ownership, surveillance, digital exclusion, and unequal power relations, demonstrating how these issues influence both AI governance and everyday working conditions.

Participants will gain practical insights into designing AI systems that recognise invisible labour, protect vulnerable communities, and promote equitable data governance. The session combines empirical evidence from African contexts with policy and design recommendations relevant to researchers, practitioners, policymakers, and technology developers interested in responsible AI