Artificial intelligence is already widely used in education in at least the following areas: (1) intelligent tutoring and personalized learning; (2) adaptive assessment of learning outcomes; (3) virtual teachers and teaching assistants; (4) intelligent classrooms and learning environments; and (5) learning diagnostics and academic prediction. Advances in artificial intelligence will continue to drive innovation and improvement in education, providing better learning and teaching experiences. Artificial intelligence is also driving developments in mathematical research. For example, machine learning is used to help mathematicians discover patterns and make conjectures in pure mathematics, such as in the algebraic and geometric structure of knots, and predicate the combinatorial invariance conjecture for symmetric groups, even solving geometric problems in IMO with the DeepMind geometric reasoning model (AlphaGeometry). But AI has not been connected to mathematical problem-posing.

Mathematical problem-posing is a complex intellectual activity that trains students' mathematical creativity and critical thinking. In the age of artificial intelligence, we need to consider how to use generative AI to engage in mathematical problem-posing activities and pose valuable mathematical problems. Therefore, the blueprint of this research is to explore the mathematical problems posed by generative AI. Applying the same mathematical problem-posing task, a paper and pencil test is used for the participants, and some prompts are used for the generative AI. Then, using textual analysis, we analyze and compare the similarities and differences between the problems posed by each. The results demonstrate that the problem-posing products of humans and AI are different, and that there are differences in the number, solvability, clarity, and complexity of the mathematical problems posed by them. The mathematical problems posed by generative AI have unknown characteristics and creativity. This research will be new and imaginative.