Co-pilots in problem solving: A qualitative inquiry into AI-assisted learning in mathematics

Abstract

Artificial Intelligence (AI) has become integral to education, transforming the way students approach mathematics. This qualitative study examines the perceptions, experiences, and usage patterns of BS Applied Mathematics students in using AI tools such as ChatGPT, Gemini, Meta AI, and Cici for solving mathematical problems within the context of Technology-Enhanced Learning (TEL). Grounded in the Technology Acceptance Model (TAM), it explores five dimensions: usefulness, ease of use, attitude toward use, behavioral intention, and actual use. Data was analyzed using both phenomenological and thematic approaches, incorporating interviews, focus group discussions, and observations. Students primarily used AI tools for speed, convenience, and quick solutions under time pressure. While most preferred AI as a complement to traditional learning, they expressed concerns about overreliance, limited understanding, and reduced motivation. Nonetheless, students demonstrated digital literacy by validating AI outputs and using multimodal resources, such as YouTube tutorials, to enhance comprehension. The study advocates for balanced AI integration in education, emphasizing digital literacy, responsible use, and the development of improved AI-based platforms to strengthen multimodal and explainable learning within TEL frameworks.