Teaching medical students to learn effectively in the age of AI.

What was the educational challenge? Medical students are increasingly using artificial intelligence (AI) tools for studying, including for summarization, explanation, practice question generation, and tutoring. While these tools expand access to learning support, they also shift the cognitive work of learning by enabling processes such as synthesis to be outsourced. Instruction on study strategies, however, remains rooted in a pre-AI paradigm, creating a disconnect between how students are taught to learn and how they study. Without grounding in learning science, AI use may promote passive engagement and superficial understanding. What was the solution and how was it implemented? We implemented a Principles-First AI Learning Model that positions AI as an implementation layer for evidence-based learning strategies. The model emphasizes: (1) teaching core learning principles such as retrieval practice, spacing, and metacognition; and (2) guiding students to use AI to apply these principles. Implementation occurred within an academic success program through coaching, workshops, and integrated teaching practices. What lessons were learned and what are next steps? Students value guidance on how to use AI effectively. Effective use is not intuitive and requires explicit instruction. In our observations, high-performing students often used AI to extend cognitive effort through self-testing and reasoning rather than replace it. AI may also offer potential opportunities for accessibility and scalability, particularly when grounded in learning science. Future work will evaluate the model's impact on learning outcomes, study behaviors, and AI use. Ongoing efforts will expand faculty development and explore adaptation across diverse educational contexts to support scalable and equitable learning.