Four US States Now Force Districts to Write an AI Policy
Idaho, Maryland, Oklahoma and Virginia now require every district to adopt a state-aligned AI policy, with Maryland's deadline landing this autumn.
The mandate wave keeps growing
Ohio's compulsory AI-policy deadline came and went this year with patchy compliance. Now four more states have decided that asking districts nicely is not enough. Idaho, Maryland, Oklahoma and Virginia have all passed 2026 laws requiring every public school system to adopt a locally written AI policy that aligns with statewide guidance, according to K-12 Dive. This is no longer a pilot idea confined to a handful of ambitious states; it is becoming the default expectation for American public education.
The timing matters. Maryland's Artificial Intelligence Ready Schools Act, signed by Governor Wes Moore on 26 May 2026, gives local school systems 120 days from the release of state guidance to adopt a compliant policy, according to Maryland Matters. That guidance landed in the summer, which means districts are working against a fall 2026 deadline right now, as term starts.
The shift is from "should schools have an AI policy" to "which of eight mandated elements does yours cover, and who signs off on it".
What Maryland actually requires
The Maryland State Department of Education has built its framework around eight elements: student-centred use, equity and access, data privacy and security, academic integrity and assessment, professional learning, AI literacy and curriculum, and operational uses, according to reporting from Citizen Portal and WTL Governance. Districts also have to designate a named coordinator at central-office level, a non-instructional role tasked with managing what the law calls the "productive and ethical use" of AI systems. MSDE's career-connected learning lead, Richard Kinkade, is running the state-level effort, and the department has published a Local Action Planning Guide, a Classroom Companion and an AI Tool Evaluation Rubric to help districts avoid writing their policies from scratch.
Oklahoma's version, due before the 2027-28 school year, takes a narrower but sharper line: it explicitly bars AI tools from being the primary basis for grading, discipline or other high-stakes decisions about students. That is a rule with teeth, and one that other states have so far only gestured towards in soft guidance rather than statute.
Why this differs from Ohio's stalled mandate
Ohio required policies but left districts largely to interpret the requirement on their own, and compliance has been uneven months after the deadline passed. The newer wave of laws tries to close that gap by pairing the mandate with a detailed state framework, a named coordinator, evaluation tools and, in Maryland's case, a hard clock. Whether that produces better policies rather than just faster ones is the open question. A rubric and a template reduce the excuse for inaction, but they do not guarantee a district thinks hard about what its policy should actually say for a Year 9 humanities class using a chatbot for research.
The UK contrast
England has no equivalent statutory requirement forcing every school or trust to publish an AI policy tied to a national framework, coordinator role and evaluation rubric. KCSIE's 2026 safeguarding updates touch AI risk, and the DfE has issued voluntary guidance, but nothing compels a school to name an AI lead or work through a checklist of academic integrity, data privacy and equity provisions the way Maryland now does. For UK school leaders watching this from a distance, the interesting bit is not the mandate itself but the mechanism: a state department doing the guidance-writing so individual schools do not each have to draft governance from a blank page.
What to do
- If you lead a school or trust, check whether anyone holds explicit responsibility for AI policy and tool evaluation, formally or otherwise. Maryland's coordinator model is worth borrowing informally even without a legal mandate.
- Look at Oklahoma's high-stakes-decision restriction as a template clause. A written line that AI outputs cannot be the sole basis for a grade or disciplinary decision is cheap to add and closes a real gap.
- Ask whether your policy, if you have one, actually distinguishes between classroom use, assessment integrity, data privacy and operational admin uses, rather than treating "AI" as a single undifferentiated risk to manage.
What to watch
- Whether Maryland districts hit the fall 2026 deadline, and whether MSDE publishes compliance data the way it has promised.
- Oklahoma's 2027-28 deadline and how districts interpret the ban on AI-driven high-stakes decisions in practice.
- Whether more states follow with paired guidance-plus-mandate laws rather than Ohio's unsupported requirement, and whether that measurably improves policy quality rather than just paperwork volume.
Sources: 4 more states require districts to adopt AI policies (K-12 Dive), Maryland schools face fall 2026 deadline for AI policy, MSDE releases K-12 AI guidance and planning tools for districts, Maryland AI Ready Schools Act Clears Senate (WTL Governance), Maryland school districts face fall deadline to set AI policies (Maryland Matters)