Generative AI in Programming Education: Evaluating PTIK Students’ Ability to Detect, Explain, and Correct Code Errors

Generative Artificial Intelligence (GenAI) has transformed programming education by enabling students to generate, modify, explain, and debug AI-assisted code. However, AI-generated code is not always accurate, while improved task performance does not necessarily reflect programming understanding. This study examines students’ competencies in evaluating GenAI-generated code, particularly error detection, explanation, and correction. A qualitative library research approach was employed by synthesizing recent literature on GenAI, programming education, debugging, critical thinking, responsible AI use, and programming assessment. The literature was analyzed thematically through familiarization, coding, categorization, comparison, and synthesis. The findings identified five major themes: changing student interaction with GenAI, the high but imperfect capability of AI-generated code, challenges in evaluation and debugging, the relationship between AI-assisted programming and learning, and the need for responsible AI use and redesigned assessment. The study emphasizes that programming competence should include verification, explanation, debugging, and correction of AI-generated solutions. This perspective supports AI-integrated programming education and assessment while preserving critical reasoning and human responsibility.