With students using tools like Copilot and ChatGPT for nearly every programming assignment, it is becoming increasingly difficult to gauge who actually understands the underlying logic and who is simply auto-completing their way through a script. While preparing students for industry means acknowledging that these tools are standard practice, over-reliance risks masking major gaps in their foundational problem-solving skills. What strategies have you found successful for modifying exams or lab assessments to verify that students can still think through an algorithm independently, without forcing you to completely ban AI from your classroom?
