This work aims to present how ChatGPT can contribute to the study of some topics in Mathematics and Mathematics Education. It is exploratory research with a qualitative approach, seeking to understand the problem in a more detailed way and making it more explicit. A historical overview of the evolution of artificial intelligence was elaborated, the structure of ChatGPT was investigated, addressing its potentialities and limitations, and its use in the educational context was analyzed, with a specific focus on Mathematics Education. The model's responses and its ability to solve some mathematics content were examined, highlighting both the potentialities and the limitations, exploring strategies for the creation of prompts in the use of generative artificial intelligence tools. The results indicate that ChatGPT has the potential to support mathematics teaching and learning processes, especially due to its ability to structure explanations and respond quickly to requests. However, limitations related to mathematical accuracy, the occurrence of hallucinations, and difficulties in contextual interpretation were observed. These results are exploratory in nature and reinforce the need for critical thinking and pedagogical validation of the responses produced.