Assessment of Coding Skills and Programming Knowledge in the Age of Generative AI

The rapid emergence of artificial intelligence (AI) and generative AI tools poses a significant threat to the validity of traditional programming assessments. As the boundary between authored and AI-generated code becomes increasingly indiscernible, long-standing assessment models centered on output correctness and code submission are at risk of obsolescence. Despite its urgency, prior work has largely concentrated on detection, with limited emphasis on reimagining assessment design. This chapter addresses that gap by proposing strategies for assessing programming proficiencies in an AI-mediated context. Its objective is to help educators move beyond surveillance-based models and adopt approaches that emphasize uniquely human cognitive capacities. These pedagogical strategies advance the field by shifting the discourse from reactive prevention to proactive, pedagogically aligned assessment design. In doing so, the chapter affirms that the future of programming assessment is not about resisting AI but about designing systems that assess human thinking over code output.