Many school districts still lack clear guidance for teachers and students.
by Robbie Sequeira for state line
With many students and educators already using widely available artificial intelligence tools, state legislatures and school districts are trying to catch up on AI policy.
In Maryland, for example, the policy of using AI in K-12 schools is “all over the place,” said Democratic state Sen. Katie Fry Hester.
He said some school districts are encouraging the use of AI, while others are restricting its use or, for Hester, the worst-case scenario, there is little or no policy guidance.
“What we heard over and over again was that teachers felt like they had to operate the AI completely on their own,” Hester said.
Hester said the most important thing for lawmakers is AI literacy, which is the purpose of a new bill she introduced and signed into law in May. We need an AI coordinator in each school system, statewide AI professional development for teachers, and AI literacy as a component of career readiness and computer science standards for K-12 students. It also requires the state Department of Education to provide some guidance on AI.
Many other states are also developing AI policies for schools. Lawmakers have introduced more than 134 bills in 31 states this year related to AI in education, focusing on data privacy, classroom restrictions, literacy, training, and more, according to government affiliate Multistate.
According to research from the Center for Democracy & Technology, the majority of teachers (85%) reported using AI in the classroom in the 2024-25 school year, and 86% of students said they used AI for personal or school-related reasons. However, only about half of teachers and students reported receiving training or information about AI from someone at their school, and very few received training or information about the risks of using AI.
Noel Ellerson-Ng, chief advocacy and governance officer at the School Superintendents Association, said the rollout of ChatGPT in 2022 marks a turning point for schools. “AI was not something we could manage,” says Ellerson-Ng. “The moment students had access to it, it was in the classroom.”
Her association does not take a position on state AI bills or policies. But he said boroughs are trying to avoid haphazard and reactive policies, such as New York City’s short-term ban on ChatGPT in 2022, due to concerns about fraud.
Some states are making progress in laying the groundwork for K-12 AI policies.
Ohio has set a July 1 deadline for all school districts, community schools, and STEM schools to adopt AI usage policies. The state’s model policy recommends that school districts address student and faculty use, privacy, ethical use, teacher-specific use, vendor contracts, third-party AI tools, and student assessments.
A new Idaho law signed in March requires local school districts and charter schools to develop local policies regarding the use of AI in K-12 schools, mandates state standards for AI literacy and training, and ensures that AI does not “replace or displace human teachers.”
An Oklahoma law enacted last month requires AI tools to be age-appropriate and requires teachers to review AI-generated materials before using them in the classroom. Parents can also opt out of the use of AI tools for their children. The legislation also directs the state Department of Education to develop AI instruction and requires local school boards to develop policies by the 2027-28 school year.
However, despite the large number of vendors selling AI products to schools, there is growing skepticism about AI in the classroom. This has led more states and school districts to implement cell phone bans and reconsider their reliance on laptops, following similar backlash over the academic and mental health effects of social media and digital technology on students.
A Center for Democracy & Technology survey found that half of students felt that using AI in class made them feel less connected to their teachers, and 70% of teachers said they were concerned that students’ use of AI was preventing them from learning important skills.
Schools need to weigh the benefits of introducing AI tools in the classroom against the impact on students’ privacy, mental health and social skills, said Sue Torts, director of outreach at Common Sense Media, a nonprofit advocacy group focused on technology and its impact on children and families.
Toots said schools may be “the only mandated safe place” where students can learn how to use and access emerging technology. But she and other education experts believe districts need to step up their scrutiny of products.
The global market for AI products in K-12 schools will be worth about $391.2 million in 2024 and could rise to more than $9 billion by 2034, according to market research firm market.us. This includes AI products for tutoring, personalized learning, automated grading, lesson planning, and administrative tasks.
“When I talk about AI literacy, it’s not about how to use AI; it’s about understanding how AI is built,” Thotz says. “Why is it being made? Who benefits from this?”
“Give the kids tools”
New York State Assemblyman Robert Carroll said he uses artificial intelligence in his own work and recognizes its value. Having struggled with dyslexia as a child, he believes technology can also help students with disabilities.
But he also wants to eliminate AI from most classroom instruction, from kindergarten through middle school. He said students first need to learn basic subject matter combined with critical thinking and then use tools that can support them.
Carroll, a Democrat, has introduced a bill that would ban the use of most AI in kindergarten through eighth-grade classrooms, except for diagnostic testing and support for students with disabilities.
“It is essential that all children acquire strong foundational skills, especially in reading, writing and numeracy, and AI appears to be in a unique position to potentially undermine that,” he said. “There’s a difference between giving tools to an adult and giving tools to a child who hasn’t yet mastered the skill.”
Most bills seeking to limit AI have chosen to focus on age limits, parental opt-outs, monitoring, and bans on using AI to replace teachers, rather than outright bans.
Related | All the ways AI is already ruining you
Florida’s AI Bill of Rights proposal this year would have included statewide restrictions on students’ access to AI educational tools until sixth grade, with the exceptions of supervised use by school staff, translation support for English language learners, and accommodations for people with disabilities. It passed in the Senate with an overwhelming majority of 37 to 1, but was rejected in the House of Representatives.
A new law in Connecticut adds computer science, including AI and emerging technologies, to the required curriculum in public schools. Connecticut lawmakers have failed to pass a bill aimed at preventing public school educators from being “replaced” by AI in 2025.
Sophia Romy, general manager of GenAI Studio, an initiative at the College Board, a nonprofit that administers the Advanced Placement curriculum and SAT tests for high schools, that studies how students and educators use generative AI, said she is concerned that only about one in five school districts allow students to use generative AI and have formal policies governing its use.
Romy said College Board research shows that many students are concerned about relying too much on AI, and that adults need to provide clearer guidance when using AI tools for brainstorming, correction, and tutoring crosses ethical lines and constitutes cheating.
“Students are much more aware of the risks of AI than the headlines suggest.”
Like aviation in 1905
Jason Corey, director of the Center for Academic Innovation at Maria University in Albany, New York, said policy discussions need to move beyond whether schools are “for” or “against” the use of AI.
“The better question is: What kind of AI use is monitored, age-appropriate, transparent, and connected to real-world learning,” Corey said. He said schools need guardrails around privacy, student data, bias, teacher training and equity of access, but also permission to “experiment responsibly.”
Ellerson Ng of the School Superintendents Association said superintendents see AI as part of a larger umbrella of disruptive technologies in schools that have evolved from calculators to laptops to cell phones. The lesson, she says, is that overreactive policies rarely work. He also said schools should address AI as part of a broader technology policy, rather than a separate policy.
“I don’t have a calculator policy. Why do I have an AI policy?” she said, explaining how some district leaders are thinking about the issue. “I have a technology policy.”
Related | Americans don’t want a data center in their backyard
In the past, technology such as cell phones and laptops often allowed adults to control when students had access, Ellerson-Ng said. With AI apps and platforms, many students had access to the tools before teachers, principals, or state officials even noticed.
That makes it difficult to ban, she says. Schools can block school-owned devices and tools on their networks, but “the only thing that stands apart from the presence of social media and AI in schools is a personal device.”
Justin Reich, associate professor of digital media at the Massachusetts Institute of Technology (MIT), said the uncertainty surrounding AI should make policymakers wary of declaring best practices too soon.
“Writing a guide to AI in 2026 is like writing a guide to aviation in 1905, before airlines, airports, or even commercial flight,” Reich said, noting that states are now trying to regulate AI in the classroom while the field is still very unstable.
“If you were to take any of the AI literacy documents, the AI readiness documents, or even the moratorium documents and run them against a checklist, you would see a lot of boxes in the “We’re making this up” column, but not as many boxes in the “We have evidence” column, Reich said.
Reich said state legislatures and school districts should be honest that they don’t know what they’re doing, rely on limited expert input, and that policies are subject to change based on new information.
“Lawmakers are going to have to be honest about what they’re proposing now that could be completely outdated in two years.”
