Artificial Intelligence in Physics Classrooms: Comparative Perspectives on ChatGPT and DeepSeek as Learning Supports

This study presents a conceptual comparative analysis of two advanced AI-driven platforms, ChatGPT and DeepSeek, within the context of physics education. It explores how these systems can serve as complementary learning supports that enhance conceptual understanding, inquiry, and problem-solving in classroom settings. ChatGPT’s dialogic and adaptive interaction style fosters exploratory learning, metacognitive reflection, and conceptual clarification, aligning well with constructivist and inquiry-based pedagogies. In contrast, DeepSeek’s precision-oriented architecture and iterative refinement processes lend themselves to structured problem-solving and cognitive efficiency, supporting students in procedural and computational tasks. Together, these affordances illustrate how distinct AI designs can address different dimensions of physics learning. The analysis also examines emerging ethical and pedagogical challenges, including issues of academic integrity, cognitive dependency, and algorithmic bias, underscoring the importance of teacher mediation and institutional regulation. The study concludes that, when thoughtfully integrated, AI tools like ChatGPT and DeepSeek can enrich physics education by balancing conceptual exploration with procedural rigor, while emphasizing the need for ethical literacy and transparent frameworks in their classroom use.