This chapter follows the debate of data work, without which AI-based technologies would not exist. It introduces concepts that capture bias in datasets and algorithms from a feminist data ‘ethics of care’ approach and discusses approaches to avoid or countervail bias. The chapter then turns to work and the question how to make AI-based technologies work in practice. Examples, most of them from IT development and health care, help understand the centrality of care, trust, and human–algorithm collaborations, what is often called ‘the human in the loop’, as key elements determining the usefulness of AI-based systems and tools to work. Trust and care are central to data and algorithmic stewardship, which will need to be mobilized in relation to AI and machine learning if we are going to achieve gender justice in future work processes.