Maryland’s Artificial Intelligence Ready Schools Act took effect on June 1, 2026. The law requires statewide guidance for school communities, establishing a framework for classroom technology rather than mandating a particular chatbot. Its requirements appear in legislation published by the Maryland General Assembly.
Utah is taking a different route through a statewide Gemini for Education partnership. California offers another comparison: adaptable policy guidance. For teachers, parents, and school leaders, these approaches raise a shared question: how quickly should schools introduce AI while building the safeguards and skills needed to use it well?
Maryland Makes Oversight a Requirement
Maryland’s law directs the Maryland State Department of Education to publish guidance addressing responsible use, privacy, accessibility, and educational needs. Local school systems must adopt aligned policies within 120 days after that guidance is released. That deadline runs from publication, rather than the law’s effective date.
The legislation also requires evaluation tools for reviewing AI products and professional development for educators and school leaders. Training must cover classroom applications, security, academic integrity, and avoiding overdependence. This approach treats readiness as an institutional responsibility, with standards intended to guide decisions across districts.
Utah Expands Access, but Districts Choose
The Utah State Board of Education announced Gemini for Education for the 2026–2027 academic year through a partnership with Google. However, local education agencies retain discretion over adoption. Statewide availability therefore does not mean every classroom must use the platform.
The board describes potential uses including lesson planning, practice materials, brainstorming, and explanations of difficult concepts. Its announcement promises educator training and accompanying AI literacy instruction. It also states that conversations will not train AI models and that schools retain administrative control over their data.
California Leaves Room for Local Adaptation
The California Department of Education posted a model AI policy on June 25, 2026. Its July announcement identifies academic integrity, disclosure, privacy, educator discretion, and parent review rights among the main topics. The template is voluntary and encourages local modifications.
Compared with Maryland’s statutory duties and Utah’s platform partnership, California emphasizes a framework districts can tailor. That flexibility may suit different communities, but it also makes local policy choices especially important.
Early Adoption Needs More Than Accounts
Utah’s stated goals illustrate the argument for moving early: teachers could receive practical support while students learn to question AI outputs. Yet these are intended benefits, not proof that statewide access improves achievement. The board itself includes verifying accuracy and reliability among its learning priorities.
Maryland’s training requirements and California’s safeguards point toward a balanced response. Schools should test specific uses, explain acceptable assistance, and protect independent work. Teacher preparation deserves the same attention as software availability.
What Should Change Before 2027?
For the second semester, leaders should review classroom results before expanding access. Maryland requires AI literacy within relevant standards by June 1, 2027. Across these approaches, the useful test remains whether technology supports learning under clear, workable human oversight.