Yapay Zekâ ile Akıllı Kodlama: Eğitim Alanında Kuramsal Bir İnceleme ve Uygulama Odaklı Çerçeve

ABSTRACT AI-based coding assistants and large language models (LLMs) have recently transformed software development practices, necessitating a reconsideration of the goals, instructional strategies, and assessment approaches of programming education. “Intelligent coding with artificial intelligence” encompasses a broad range of skills, including expressing intent in natural language to generate code, debugging, writing tests, refactoring, and receiving personalized feedback throughout the learning process. Adopting a theoretical approach that does not require field research, this article examines (i) the educational potential of AI-supported coding, (ii) learning-related risks (such as overreliance, superficial learning, academic integrity concerns, security vulnerabilities, and erroneous or hallucinated outputs), (iii) concrete classroom scenarios for instructional design, and (iv) ethical and governance principles within an integrated framework. The discussion is grounded in the literature on the general capability profiles of LLMs, evaluation criteria for code generation, software engineering benchmarks, and emerging guidelines for the use of generative AI in programming education. Ultimately, the article conceptualizes “intelligent coding literacy” through the components of problem solving, conceptual understanding, verification and testing culture, security awareness, and ethical responsibility, and proposes a practical instructional and assessment roadmap for programming courses. Keywords: Artificial Intelligence, Intelligent Coding, Programming Education, Generative Artificial Intelligence, Large Language Models, Assessment-Evaluation, Academic Integrity, Code Security.