The integration of Artificial Intelligence into programming environments presents both unprecedented opportunities and underexplored challenges for software developers. WhileAI-powered coding tools such as large language model (LLM)-based code assistants havedemonstrated impressive generative capabilities, their design has largely been driven bytechnical benchmarks rather than the cognitive, social, and contextual needs of human developers. This article argues for a paradigm shift toward Human-Centered AI (HCAI) inprogramming environmentsone that prioritizes usability, transparency, trust calibration,and developer agency. Drawing on frameworks from Human-Computer Interaction (HCI),Cognitive Science, and Software Engineering, we outline the theoretical foundations, keyresearch challenges, and methodological directions for this emerging eld. We propose afour-pillar framework encompassing cognitive augmentation, transparency and explainability,developer agency, and collaborative dimensions, and identify open problems uniquely suitedto doctoral-level inquiry.
Sandeep Kumar Bollu (Wed,) studied this question.