Most U.S. K-12 Teachers Report Little Formal Guidance on AI Use, Study Finds
A national survey conducted by the Walton Family Foundation and Gallup between February 9 and March 2, 2026 found that the majority of public K‑12 teachers in the United States have received no formal instruction on how to use artificial intelligence (AI) tools in the classroom. The study, titled Teaching for Tomorrow: Closing the Expectations Gap, surveyed 2,069 teachers drawn from the RAND American Teacher Panel, a nationally representative group of educators.
Only 18 % of teachers reported receiving any type of formal guidance—such as written policies or official directives—from school administrators about AI use. In contrast, 48 % of respondents said they had received only informal guidance, and 34 % reported having received no guidance at all. The survey also asked teachers to rate how often they use AI for ten specific tasks, ranging from lesson planning to grading.
The tasks that garnered the most guidance—whether formal or informal—were lesson preparation, creating assignments and activities, making assessments or exit tickets, and modifying materials to meet student needs. For example, 45 % of teachers said they received no guidance on using AI to prepare lessons, while only 9 % reported formal guidance for that purpose. Similarly, 47 % of teachers said they had no guidance on using AI to create worksheets, assignments, projects or classroom activities, and just 8 % received formal instruction for that use.
When it comes to more sensitive or individualized applications, the lack of guidance is even more pronounced. According to the study, 58 % of teachers reported no guidance on using AI for grading and providing feedback, and 69 % reported no guidance on using AI in one‑on‑one instruction. Only 9 % of teachers said they had received formal guidance on creating assessments, quizzes or exit tickets, despite a quarter of teachers in the 2025 survey reporting monthly use of AI for that task.
The report notes that most of the guidance teachers do receive is informal. “For many teachers, decisions about if, when and how to use AI are being made individually, in the absence of clear institutional expectations,” the study states. Rob, a teacher in the Tempe Union High School District who requested anonymity, echoed this sentiment. He said the professional development he has received is “tool‑focused—here’s a platform, here’s how to use it—rather than built around clear expectations for teachers or students.” Rob added that the lack of explicit guidance creates a “gray area” where teachers must judge acceptable use on their own.
Some educators are stepping in to fill the gap. Amber Gould, an English teacher in Glendale Union School District, has led AI‑related training and discussions in her district and through the Arizona Education Association. Gould said interest in the topic has grown as AI tools become more common among teachers and students. She emphasized that conversations should focus on AI literacy, ethical use, and how to incorporate the technology without replacing critical thinking or teacher judgment.
The study also found disparities in guidance availability. Teachers in higher‑needs schools were less likely to receive guidance than those in wealthier districts, particularly for creating student materials. In wealthier districts, 59 % of teachers reported some guidance on using AI to make worksheets and assignments—10 percentage points higher than the highest‑need schools.
Arizona lawmakers introduced several bills related to AI in schools during the current session, including House Bill 4005, which would require districts to provide instruction on the ethical, moral and educational uses of AI beginning in the 2027‑28 school year. None of the proposals, however, specifically address the provision of formal AI guidance or training for K‑12 educators.
The study’s findings suggest that as AI tools become increasingly common in classrooms, many teachers are navigating the technology largely on their own, without formal policies or consistent guidance from schools and administrators. The lack of clear expectations could lead to uneven AI use and potential risks related to equity, privacy, and educational integrity.
The research was made possible through grant funding from the Arizona Local News Foundation’s Arizona Community Collaborative Fund.