This chapter examines the application of machine learning in legal contexts, tracing developments from early AI systems to contemporary natural language processing technologies. Machine learning techniques may address access to justice challenges through automated legal analysis and case outcome prediction. This chapter discusses different approaches to extracting useful information from legal documents and analyzes research on predicting legal decisions. A central focus is the tension between powerful but opaque machine learning models and the legal system’s requirement for transparent, explainable reasoning. Key challenges include the difficulty of representing legal cases effectively for machine learning, the limited explanatory capacity of high-performing neural networks, and the gap between how judges and litigants describe facts. While machine learning shows promise for legal applications, significant obstacles remain in developing systems suitable for actual legal practice where explanation and justification are essential.