Guide to Integrating Generative AI for Deeper Math Learning

An initial exploration of how to harness generative AI's (GenAI) power to enhance, not replace, the cognitive lift and meaningful learning in K-12 math classrooms.

Watch the accompanying webinar.

Developed in collaboration with Student Achievement Partners , this document explores practical strategies for using AI to support students’ productive struggle while maintaining the essential cognitive work that drives learning. With an understanding that both the technology and its educational applications are rapidly evolving, we offer these insights as a conversation starter and support as educators continue to wrestle with how GenAI can serve instructional needs and how content and pedagogy intersect with this evolving technology.

Key Highlights:

• The guide distinguishes between productive struggle (which enhances learning) and counterproductive struggle (which hinders learning) in mathematics education

• GenAI should support, not replace, productive struggle and meaningful engagement with mathematical concepts

• For computational tasks, specialized math-specific AI tools (like Khanmigo or Desmos) are strongly recommended over general-purpose LLMs

• At the elementary level, GenAI tools should primarily be used by teachers for lesson planning and preparation, not by students directly

• The document provides practical use cases for both students and teachers with detailed examples, instructional benefits, and caution points

Productive and Counterproductive Struggle in Math Classrooms

• Productive struggle builds conceptual understanding, offers multiple solution pathways

• Counterproductive struggle focuses on procedural skills and memorization without conceptual understanding

• Detailed chart contrasts productive vs. counterproductive struggle across four components of high-quality math instruction

Enhancing Math Instruction with GenAI

• General-purpose LLMs (ChatGPT, Claude) have significant limitations in calculation reliability and accuracy

• Specialized math tools receive additional fine-tuning for mathematical concepts

• Key principles include ensuring tools support (not circumvent) productive struggle

• AI tool categories and their optimal uses

Specialized AI Tools for Mathematics

• Examples include Khanmigo, Mathway, Photomath, Gauthmath, Snorkl, Mathos AI, Wolfram Alpha, Desmos, and Julius AI

• These tools are better suited for computational tasks than general-purpose LLMs

Key Principles for GenAI Integration

When leveraging AI tools in math instruction, the following rules apply:

• Tools should support, not circumvent, productive struggle for students.

• Integration should enhance, not replace, proven instructional practices.

• Usage should align with students’ developmental readiness and mathematical learning goals.

• GenAI should augment educators’ pedagogical expertise, content knowledge, and knowledge of students.

• LLMs should be used for language-based tasks like conceptual exploration, planning, and generating examples 

• Specialized math-specific tools should be used for tasks requiring c

Note About GenAI Use for Elementary Teachers

• Elementary students are still developing critical cognitive skills needed to evaluate AI content

• Until more evidence supports direct student interaction, elementary teachers should restrict GenAI use to their own workflows

• Many GenAI tools have age restrictions that prevent direct elementary student use

Strategies for Using GenAI to Support Deeper Math Learning

• Student use cases outlines three specific ways students can leverage specialized AI tools to enhance their learning while maintaining productive cognitive engagement

  • AI Tutoring: Uses specialized tools to help students think through problems, offering private support without judgment

  • Error Analysis and Feedback: Identifies specific areas where understanding breaks down, providing safe spaces for students to ask questions

  • Procedural Skills Practice: Creates personalized practice problems aligned with curriculum, allowing students to customize content to their interests

• Teacher use cases presents six strategic applications for educators to incorporate GenAI into their instructional practices while preserving the essential mathematical thinking processes

  • Creating Lessons and Materials: Helps develop lesson plans, instructional content, and targeted interventions aligned to learning objectives

  • Generating Authentic Practice Problems: Creates contextual problems incorporating real-world scenarios and student interests

  • Providing Differentiation: Develops modified assignments with multiple entry points for diverse learners

  • Breaking Down Tasks and Concepts: Unpacks challenging content by identifying prerequisite skills and common errors

  • Analyzing Student Data: Identifies trends and action items while protecting student privacy

  • Creating Visual Representations: Generates graphs, charts, and interactive tools to illustrate mathematical concepts

• Each use case includes detailed descriptions, instructional benefits, and caution points

Looking Ahead

As AI tools and their applications in education continue to evolve, this resource will grow and adapt based on emerging research and classroom experience. We invite educators to share their experiences implementing these strategies, provide feedback on what's working, and contribute insights about the impact on student learning. Your classroom expertise will help shape future versions of this resource and deepen our collective understanding of effective AI integration in math instruction.

Please reach out to us with your observations and suggestions here.

Want to partner with AI for Education at your school or district? LEARN HOW