MAIA is the abbreviation for Generative AI in Mathematics Teaching: a tool for problem-solving and a resource for mathematical thinking. The project is being conducted in collaboration with upper secondary schools.

“We are particularly interested in what happens during the problem-solving process– how students think, reason and learn,” says Pernilla Josefsson, senior lecturer in Media Technology at the School of Natural Sciences, Technology and Environmental Studies.

The project is in its initial stages and takes the school’s normal activities as its basis.

“The questions we are working with originate from the teachers’ own practices. They see an increasing need to understand how generative AI is changing what mathematical comprehension entails.”

And this is a significant need; mathematics is a subject in which many students struggle to fulfil the learning objectives. According to statistics from the Swedish National Agency for Education, in the spring of 2025 mathematics was the subject in which the highest number of students failed to achieve a pass grade, among those who did not meet the entry requirements for national programmes at upper secondary school. This makes questions about how students arrive at mathematical solutions – and AI’s role in this process – particularly pressing.

From quick answers to mathematical thinking

Generative AI is often used as a tool for the rapid production of answers. One key question in the project is whether this technology can also help students gain deeper understanding.

“It’s not just about producing an answer, but about supporting the students’ understanding of concepts and mathematical reasoning.”

The researchers also want to investigate when AI could obstruct learning, such as if pupils rely too heavily on ready-made solutions. As the project is still at an early stage, there no conclusive results, but the researchers highlight the importance of studying how AI is used in teaching practices.

Will provide classroom support

The project is being conducted in collaboration with Widerströmska School in Huddinge and NTI School in Stockholm. The practice-oriented approach is vital for its aim of generating knowledge that is directly applicable in teaching.

“The issue is not whether to ban or uncritically embrace technology. Rather, it is understanding how it can be used – and when it should not be used – with a focus on enhancing student learning.”

The ultimate aim is to provide models and methodologies that can help teachers plan, deliver and evaluate AI-supported maths activities in an educationally sound and well-considered manner.”