This paper presents a pedagogically grounded and adaptable framework for integrating artificial intelligence (AI) into engineering education. Unlike existing approaches that focus on tool usage, this work formalizes how AI can be embedded within structured learning processes to enhance critical thinking without compromising foundational knowledge. The framework integrates scaffolded instructional design principles with AI chatbot-based reflective interaction, combining the pedagogical rigor of structured stepwise learning with dynamic feedback mechanisms. A Linear Algebra module on Gaussian elimination is used as a validation case study to demonstrate the implementation. The framework introduces an adaptation layer that ensures robustness to rapidly evolving AI capabilities while maintaining alignment with engineering learning objectives. Results show improved procedural accuracy, deeper conceptual engagement, and increased student ability to critically evaluate AI-generated responses. The study contributes a generalizable model for AI-assisted learning applicable across engineering disciplines.

Keywords

Artificial Intelligence in Education, Guided Problem-Solving, Engineering Pedagogy, Linear Algebra Instruction, Adaptive Learning Design