The Development of a Socratic Facilitation Layer With Generative AI in IS Database Education

Students in IS courses frequently reproduce procedures without understanding the underlying logic, exhibiting procedural fluency without conceptual depth (Anderson & Krathwohl, 2001). The arrival of generative AI tools amplifies this risk, as students can now produce technically correct outputs without engaging the conceptual reasoning the assignment was designed to develop. This paper proposes an alternative role for generative AI in IS education, as a Socratic facilitation layer rather than a content delivery tool. The paper describes a designed scaffolded inquiry exercise in entity relationship modeling, deployed within a faculty-authored web tool that loads an instructor-designed system prompt and constrains the AI to respond only with questions. The student is positioned as the reasoning agent and the AI is positioned as the questioning instrument. The paper presents the instructional architecture, a six-step pedagogical algorithm for ERD construction, design principles, and an exploratory pilot plan for analyzing student-AI interaction transcripts as both course deliverable and research instrument. Contributions include a design artifact, a set of design principles for AI-facilitated IS pedagogy, and a framework for distinguishing tool-as-answer-source from tool-as-facilitation-layer in generative AI educational deployment.