Generative AI has fundamentally changed how computer programming is taught in data science. Students now routinely rely on AI tools to generate code, explain concepts, and debug errors, raising important questions about which programming skills remain essential and how instructors should assess them. Rather than resisting these tools, we have had the opportunity to redesign programming instruction to leverage AI while ensuring students develop a strong conceptual foundation. AI enables students to troubleshoot more efficiently, allowing classroom time to shift toward more ambitious, real-world projects. At the same time, we have seen a need to reinforce core programming concepts through assessments that minimize AI assistance, such as handwritten or paper-based coding exercises that require students to demonstrate their understanding of syntax, logic, and algorithmic reasoning.
