How STEM Teachers Can Use GenAI for Real Instructional Impact 

April 24, 2026

Generative artificial intelligence (GenAI) is reshaping what’s possible in STEM classrooms, but many teachers are only scratching the surface of what it can do. WestEd’s recent Leading Together webinar explores findings from two studies conducted as part of the AmplifyGAIN AI Research and Development Center, shedding light on how math and science teachers are currently using GenAI, what’s getting in the way, and what kinds of support can help them move toward more transformative, student-centered use. 

Led by Dr. Ann Edwards, Senior Director of Mathematics Education at WestEd; Sarah Rainier (formerly Nielsen), a research associate on the mathematics team; and Dr. Drew Nucci, a research associate on the mathematics team, the session offered research-grounded insights for educators and leaders looking to expand meaningful AI use in STEM instruction. 

How Are Teachers Using GenAI Now?

A nationwide survey of approximately 1,000 K–12 math and science teachers found that about 85 percent have very limited GenAI use, meaning either they’ve never heard of it, they’ve tried it and opted not to continue, or they’ve heard of tools but haven’t yet used them. Among those who do use GenAI for instructional tasks, the most common uses are creating materials and planning lessons (76%) and generating assignments or assessments (61%). Only 22 percent of teachers have received any formal professional learning on AI, and just 5 percent work in schools or districts with formalized AI use policies. 

To make sense of how teachers are applying these tools, the research team used a three-part framework to categorize use cases: substitutive, amplified, and transformative. 

• Substitutive use does not provide better learning opportunities for students—for example, using GenAI to generate a worksheet or build a PowerPoint presentation. 
• Amplified use does lead to better learning opportunities, but those opportunities could have been achieved with other technologies—for example, using GenAI to adjust the reading levels of texts for students. 
• Transformative use leads to better learning opportunities that would not have been possible without GenAI—for example, using GenAI to generate personalized extension problems for individual students in real time based on the questions and interests they show during a lesson. 

Across the interview study with 15 focal teachers, planning was by far the most common category of GenAI use. What stood out, however, was that for instruction and professional learning specifically, the majority of use cases were transformative in nature, 

Three Factors That Shape How Teachers Use GenAI 

Through in-depth interviews with teachers and their colleagues, three key themes emerged that distinguish teachers who use GenAI in transformative ways from those who don’t use it at all or stick to basic tasks: policies, resources, and messaging. 

• Policies: Teachers with limited AI use were often in environments with no clear technology policies or with restrictive policies and little professional development. In contrast, teachers with more advanced, student-centered use worked in environments with clear instructional technology policies that actively encouraged experimentation and innovation, along with ongoing collaborative professional development structures. 
• Resources: Limited AI use often correlated with limited resources, including schools without reliable internet access. More transformative use was found in contexts with access to a range of AI tools, including student-facing tools, and where teachers were given time and financial support to participate in professional learning. 
• Messaging: In low-use contexts, messaging about AI was often vague, with conversations dominated by concerns about student cheating rather than deeper questions about what it means to prepare students for the world. In high-use contexts, AI was tied explicitly to instructional goals, leaders were transparent about their own experimentation, and collaboration and innovation were treated as shared values. 

The Supports Teachers Need 

When asked directly, teachers were clear about the kind of professional learning that would make a difference. They want PL that is grounded in their specific content area and curriculum, collaborative so they can learn from one another, hands-on and exploratory, and sustained over the course of a school year rather than delivered as a one-time workshop. 

Advanced AI users offered an additional insight for designing entry-level professional learning: Start by showing skeptical teachers how GenAI can address their most pressing pain points. Once they see what AI technology can do, that momentum can be channeled toward instructional improvement. From there, giving teachers structured time to explore and experiment together is key. 

As Dr. Ann Edwards noted during the session, the combination of “structures supporting teacher collaboration, robust professional learning alongside clear AI use policies and explicit guidance on adopted tools, all taken together can support STEM teachers to use AI toward building improved learning opportunities.” 

WestEd piloted one possible approach this past summer: a program in which teachers learned to build their own AI chatbots tailored to their instructional goals. Teachers reported learning about AI through hands-on experimentation and, importantly, gaining new ideas for designing more engaging and meaningful tasks for students.

Use GenAI for Real Instructional Impact 

Watch the full webinar, and view other webinars in the Leading Together series. Explore WestEd’s work in both Mathematics and AI and fill out our Work With Us form at the top of this page to get in touch with a WestEd expert today!