Integration of generative AI in creative mathematics learning: A correlational analysis between the quality of prompts and the logical reasoning of student teachers in the education degree programme

The presence of Generative Artificial Intelligence (GenAI) based on Large Language Models (LLMs) such as ChatGPT, Gemini, and Claude has triggered a massive paradigm shift in mathematics education. GenAI no longer functions as an information search engine (Luo et al., 2025); it has evolved into a cognitive partner capable of performing complex mathematical reasoning. The integration of GenAI into creative mathematics learning is particularly important because it enables students to explore multiple solution strategies, develop original mathematical ideas, and refine their reasoning through continuous interaction and feedback. As creativity in mathematics requires both divergent thinking and logical justification, GenAI provides opportunities for students to articulate, evaluate, and improve their mathematical arguments in real time. Various recent studies show that the use of artificial intelligence in STEM education can enhance student engagement and offer personalised learning (Huang et al., 2025; Nikhil, 2025). However, the transition from conventional learning to AI-based interaction presents new pedagogical challenges, particularly regarding how students communicate with machines to structure their mathematical thinking.