Artificial intelligence increasingly supports site analysis, massing generation, processing spatial data, and estimating design consequences. Yet its contribution to understanding the relationship between context and form in early architectural design education remains unclear. Such outputs therefore require educational mediation and critical student interpretation. Producing more data or alternatives does not demonstrate that students understand how site evidence, design criteria, and massing decisions relate. This article develops a conceptual framework for classifying artificial intelligence applications in site analysis and massing according to their educational functions. It asks how these applications can become strategies for improving students’ understanding of context and early design decisions. The study conducts a targeted, structured review of Persian and English literature on computational site analysis, machine learning, massing generation, and architectural design education, followed by analytical and argumentative synthesis. No field study, interview, or student survey was conducted. The findings identify five connected educational functions: reading and organizing site evidence; interpreting relationships and formulating criteria; generating massing alternatives; evaluating consequences; and engaging in dialogue and revision. The framework relates each technical capability to students’ cognitive activity and the instructor’s mediating role. It argues that artificial intelligence has educational value when students can trace and critically examine the path from data to decision. The framework offers a conceptual basis for integrating technical operations with design learning, while its educational effectiveness remains to be assessed through future empirical research.