State leaders can shape responsible AI use in schools, report finds
Francesco Carta fotografo via Getty Images
Two states made recent strides to go beyond suggesting safe AI policies for schools by requiring them to develop their own policies and training programs.
As the use of artificial intelligence becomes increasingly common in schools, state leaders have the opportunity to shape how education leaders implement AI policies and guidance for education officials, according to a new report.
The use of AI in education has proliferated across schools since 2022, according to a survey released last week by research organization RAND Corporation. The survey includes responses from more than 16,000 students, parents, teachers, principals and school district leaders, across eight nationally representative groups. Survey data was collected during the 2024-2025 school year.
Indeed, 54% of students said they used AI in some way for schoolwork, which is an increase of more than 15 percentage points compared with a separate survey from last year, according to the report.
Survey data also showed growth in teachers’ use of AI to assist with instructional planning or teaching. In spring 2025, 53% of English language arts, math and science teachers reported leveraging AI for those purposes, which represents an increase of more than 25 percentage points from another RAND survey conducted the prior year.
But as students and teachers increasingly turn to AI, data suggests that the development of AI policy, guidance and training lags behind the tech’s adoption.
“Ground rules really need to be there because AI can be used as a very powerful complement to learning,” said Christopher Doss, senior economist at RAND and professor of public policy at the RAND School of Public Policy. “Schools also have the responsibility or the need to teach people how to become critical consumers of AI since it’s going to become part of their everyday life.”
Without clear AI rules and standards about how the technology can be used appropriately in education, confusion among students and educators could increase, the report stated.
Indeed, less than 40% of teachers said their school or district had policy or guidance regarding AI and academic integrity, and the majority of those respondents said existing policies were “limited rather than clear and comprehensive,” according to the report.
Feedback from survey participants indicates that school leaders are looking to state officials for guidance on how to develop and implement AI policies. State-level education agencies, for example, can offer model language and sample policies as stepping stones for schools and school districts looking to craft their own rules, Doss said.
He pointed to Ohio, where lawmakers recently passed a bill in August requiring K-12 school districts in the state to adopt an AI use policy by July 1, 2026. Ohio is reportedly the first state to do so, despite nearly 30 states having some level of AI guidance for schools.
Under Ohio’s bill, the state’s Department of Education and Workforce is also required to create a model policy by the end of the year that addresses appropriate use cases for AI by students and staff and can be adopted by schools and districts.
In Tennessee, lawmakers are considering a bill that would require the state Department of Education to provide grade 6-12 teachers with a free professional development course on the use of AI in the classroom. If passed, teachers would be required to complete at least one department-approved training course by Aug. 1, 2027, or within two years of obtaining their teaching license.
While nearly 30 states already offer guidance on AI usage in schools, with suggestions on how to manage school data privacy and security, account for bias in AI tools and identify cheating, “there’s nothing binding about them,” Doss said.
More states are moving toward requiring AI policies and training among schools and school districts as people realize that “AI is not going anywhere,” he said.