Learner autonomy has generally been considered to play a central role in data-driven learning (DDL), as it enables students to take responsibility throughout the learning process. This implies that learners provide input at all stages, from initial decisions as to what and how to learn, through implementation of their chosen strategies, to evaluation of the outcomes. The social context of a DDL intervention plays an important role in developing individual autonomy through teacher input and mediation alongside interactive collaboration with other learners. In class-based DDL, teachers often have overall responsibility for macro-level decisions affecting the whole class, while learners take decisions on a micro-level that affect their own learning. Although autonomy is often claimed as a positive outcome of DDL, there is little research that sets out to test this assumption or attempts to quantify the effect of DDL on the achievement of learner autonomy. Encouraging developments include improved tools to aid autonomous learning, especially for academic writing, and the spread of open educational resources.