A growing number of states and school districts are trying to put limits on artificial intelligence in public schools, as educators, parents and researchers debate whether the technology is helping children learn or allowing them to bypass the very skills classrooms are meant to develop.
The pushback comes as the Trump administration encourages wider AI adoption and technology companies expand their education offerings.
But the movement is not a uniform effort to eliminate AI from schools. Some jurisdictions are restricting student access to chatbots, while others are establishing rules for privacy, academic integrity, teacher oversight and automated decisions about students.
Meanwhile, some school districts are treating the technology as a potential teaching aid rather than something to keep out of the classroom. In Decatur, Georgia, middle-school students have used AI to explore different ways to solve math problems, while the district has tested an AI-powered tool to help high-schoolers prepare for exams. At the state level, Alabama and Utah have incorporated AI literacy into required coursework, while Idaho, Maryland and Virginia have established frameworks or pilot programs intended to encourage AI use with safeguards.
New York and California Lead the Classroom Restrictions
New York City has adopted one of the country's most expansive restrictions on student-facing generative AI. In September, Mayor Zohran Mamdani and Schools Chancellor Kamar Samuels announced a one-year moratorium covering students from prekindergarten through eighth grade for the 2026-27 school year. AI companion chatbots are prohibited across all grades, while limited high-school pilots and AI literacy instruction remain part of the policy.
Separately, Assemblymember Robert Carroll has sponsored legislation that would prohibit most AI use in classrooms before ninth grade, with limited exceptions, for schools across the state.
The city's approach reflects concerns that students could become overly dependent on automated systems before developing essential academic and social skills. City officials say they want to ensure technology supports children's development rather than undermining it.
The restriction applies to New York City's public school system, not every school in New York state. A separate statewide proposal, Assembly Bill A9190, would prohibit most classroom AI use before ninth grade, with exceptions for certain diagnostic and special-education purposes. The proposal is distinct from the city's existing moratorium.
California is taking a different approach. The Los Angeles Unified School District announced restrictions on student use of generative AI on school-issued devices for the current school year. At the state level, California's Department of Education released guidance in January encouraging safe and effective AI use, with attention to student privacy, academic integrity, equity and human-centered instruction.
The distinction is important: California has not imposed a statewide ban on classroom AI, and Los Angeles's restrictions should not be confused with a prohibition covering every school in the state.
Four States are Building Rules Around AI
Idaho, Maryland, Oklahoma and Virginia all enacted education-specific AI measures in 2026, according to FutureEd's legislative tracker. Their laws generally seek to establish guardrails rather than remove AI from classrooms entirely.
Idaho directed its education department to develop a comprehensive framework for generative AI. The framework covers student privacy, procurement safeguards, transparency, academic integrity, AI literacy and professional development. School districts will be required to adopt aligned policies.
Maryland enacted the Artificial Intelligence Ready Schools Act. It requires statewide guidance and aligned district policies, designates AI coordinators, establishes an AI Education Collaborative and calls for evaluation of educational AI tools. It also supports teacher training and AI literacy.
Oklahoma has taken a different approach. Sen. Ally Seifried and Rep. Anthony Moore sponsored the Responsible Technology in Schools Act, which Gov. Kevin Stitt signed in May. The law requires school districts to adopt written AI policies before the 2027-28 school year, directs districts to disclose which AI tools they use and what student data those tools collect, and allows parents to opt their children out of student-facing AI without academic penalties. It also requires AI use to comply with educator-directed safeguards, reflecting an effort to regulate the technology rather than eliminate it from schools.
Virginia established an AI Innovation in Education Pilot Program to test and evaluate AI applications in schools. The law also requires guidance on safe, ethical and equitable use, with school districts expected to adopt policies aligned with state guidance.
Experts Warn That AI Could Undermine Learning
The policy debate is being reinforced by research into how generative AI affects learning.
An October 2026 expert-consensus report from researchers at the University of Minnesota's College of Education and Human Development examined 65 learning processes that generative AI could enhance or disrupt.
The researchers found particular concern about memory formation, critical thinking, writing, student motivation and persistence.
Experts also warned that AI could disrupt the social dimensions of learning, including human connection, collaboration and social-emotional development. Elementary-school students were viewed as especially vulnerable to disruption.
The report also emphasized a significant limitation: existing empirical research on AI's effects on learning remains insufficiently rigorous. The findings identify areas of concern, not proof that every use of AI harms children.
A separate June 2026 review published in Computers and Education: Artificial Intelligence analyzed 22 studies of generative AI risks in K-12 education. The authors identified potential problems involving anxiety, social isolation, dependence on AI-generated answers, reduced creative ownership and student privacy.
Researchers suggested that schools can reduce some of these risks by using AI to provide hints rather than complete answers, assessing students' work processes and teaching students how to evaluate AI-generated information.
The concerns extend beyond academic performance. In September, the Washington Post reported that educators and technology experts were raising questions about the expansion of AI tools into platforms students already use for assignments and classroom communication. Experts stressed that schools should know what information these tools collect, whether their educational benefits have been demonstrated and how children can challenge accusations of AI-assisted cheating.
Eric Klopfer, an MIT professor who co-chaired a committee studying AI in teaching, learning and research, told the Post in the September report: “It's the illusion of learning.” He said that students felt they understood the material when using the AI but struggled to apply the knowledge afterward.
The Trump Administration Is Pushing AI Adoption
The restrictions are emerging alongside a federal effort to encourage AI development and integration into education.
The Trump administration has promoted AI literacy and classroom integration as part of a broader effort to maintain American leadership in artificial intelligence. In August, the U.S. Department of Education issued guidance encouraging schools to focus on instructional value, evidence of learning benefits, educator judgment and transparency for parents.
The department said schools should be willing to abandon technology that fails to improve learning. That emphasis on evidence offers a potential point of agreement with critics who want stronger restrictions.
The administration has also worked with major technology companies on education-related initiatives. In September, the White House announced that Adobe, Amazon, Intel, Starlink, Google and Zoom had joined first lady Melania Trump's Fostering the Future Together initiative. The partners committed more than $23 million in education-related funding, tools and expertise.
AI Is Becoming a Political Issue Ahead of Midterms
The debate over AI's expansion is also reaching the campaign trail, with politicians increasingly confronting public concerns about the infrastructure powering the technology.
AI data centers require substantial electricity, land and, in some cases, water. Communities have increasingly questioned whether the economic benefits justify the costs and whether utility customers will end up paying for new power generation and grid upgrades.
In Ohio, Democrat Sherrod Brown has made data centers a line of attack against Republican Sen. Jon Husted, who has a Newsweek AI score of 3.2 ranking him as "mixed" on AI, arguing that the expansion of energy-hungry facilities threatens to raise household electricity bills.
In Texas, Gov. Greg Abbott has shifted from promoting the state as a hub for AI development toward proposing restrictions on data centers in rural communities and scaling back some tax incentives, as Democratic challenger Gina Hinojosa argues that his efforts do not go far enough to protect residents from the infrastructure's impacts.
The issue reflects a broader political tension: candidates want to attract technology investment, but voters are increasingly asking who will pay for the electricity and infrastructure it requires.
Candidates in states including Ohio, Texas, Pennsylvania and Wisconsin have been confronting growing opposition to data-center development. Wired reported this month that data centers have become a significant issue in several Senate, House and gubernatorial contests.
Newsweek’s reporters and editors used Martyn, our AI assistant, to produce this story. Learn more about Martyn here. Contact Newsweek editors on this story: Jason Lemon and Edward T. Cummins