It is September 2025 - nearly three years since ChatGPT launched - and I have a question for school leaders: does your district have a comprehensive, current AI policy? Not a hastily written acceptable use addendum from January 2023. Not a paragraph buried in the student handbook. A real policy that addresses AI use by students, teachers, and administrators, with clear guidelines, regular updates, and meaningful implementation support.
If you answered yes, you are in the minority. And that is a problem.
The State of AI Policy in K-12
Over the past year, through my work with iTeachAI Academy and my consulting with districts, I have reviewed AI policies from schools across the country. The landscape is troubling.
Some districts have no AI-specific policy at all. They are relying on general technology acceptable use policies written years before generative AI existed. These policies were designed for a world where the main technology concerns were cyberbullying, inappropriate content, and screen time. They are woefully inadequate for a world where students can generate essays, solve math problems, write code, and create images with a text prompt.
Other districts have policies that begin and end with prohibition. "Students may not use AI tools for academic work." Full stop. These policies are unenforceable, pedagogically counterproductive, and increasingly absurd as AI becomes embedded in the platforms schools have already adopted. You cannot ban AI when your district uses Google Workspace and Microsoft 365, both of which now include AI features.
A third group has well-intentioned policies that are so vague as to be useless. "Students should use AI responsibly." What does that mean? Responsible according to whom? In what contexts? The teachers trying to implement these policies have no guidance, and the students subject to them have no clarity.
Why the Gap Persists
I have thought a lot about why AI policy has been so difficult for districts to get right. Three factors stand out.
The speed problem. AI technology is evolving faster than policy processes can accommodate. By the time a district committee drafts a policy, circulates it for feedback, revises it, gets board approval, and communicates it to staff, the technology landscape has changed. Google has added new AI features to Classroom. OpenAI has released a new model. A new AI tool has gone viral with students. The policy is outdated before the ink dries.
This is a genuine challenge, and it means districts need a different approach to AI policy than they use for other technology policies. AI policies need to be living documents with regular review cycles - quarterly at minimum. They need to be principle-based rather than tool-specific, so they remain relevant as specific tools come and go.
The expertise problem. Writing effective AI policy requires understanding both the technology and the educational context. Many district leaders understand education deeply but have limited understanding of AI capabilities and limitations. Conversely, many technology advisors understand AI but lack classroom experience. Effective policy requires both perspectives at the table.
This is one reason I have added a policy module to the iTeachAI Academy course catalog. Teachers and leaders need a shared foundation of AI understanding before they can write meaningful policy together.
The consensus problem. AI in education is genuinely controversial. Parents, teachers, administrators, board members, and community stakeholders often have dramatically different views on whether and how AI should be used in schools. Building consensus takes time, and many districts have avoided the hard conversations because they know how contentious they will be.
Avoidance is not a strategy. The absence of policy is itself a policy - one that leaves every individual teacher to make their own rules, creating inconsistency and confusion for students.
What Good AI Policy Looks Like
Based on my review of dozens of policies and my work helping districts develop their own, here is what I believe effective AI policy includes:
A clear philosophical stance. The policy should articulate whether the district views AI as a tool to be integrated, a threat to be mitigated, or - most wisely - both. This framing shapes everything else.
Differentiated guidance by role. Students, teachers, and administrators use AI differently and face different risks. A single blanket policy cannot address all three. Teachers using AI for lesson planning and grading face different considerations than students using AI for assignments. Administrators using AI for scheduling and communication face different considerations still.
Context-dependent rather than tool-specific rules. Rather than listing which AI tools are allowed and which are banned - a list that will be outdated within months - effective policies establish principles and contexts. AI may be appropriate for brainstorming but not for summative assessment. AI may be appropriate for research but requires attribution. This approach survives technological change.
Data privacy protections. This is non-negotiable. Any AI tool used in schools must comply with FERPA, COPPA, and applicable state privacy laws. Policies should specify what student data can and cannot be shared with AI platforms, and who is responsible for vetting tools for compliance.
Implementation support. A policy without implementation support is a document, not a program. Effective policies include professional development requirements, classroom resources, and designated points of contact for questions and concerns.
A review schedule. The policy should specify when and how it will be reviewed and updated. Annual review is the absolute minimum. Quarterly is better.
The Cost of Inaction
I want to be direct about what is at stake. Districts without clear AI policies are exposing themselves to multiple risks.
There are legal risks. Without clear policies, teachers may inadvertently share student data with AI platforms that are not COPPA or FERPA compliant. Without clear academic integrity guidelines, disciplinary decisions are inconsistent and potentially challengeable.
There are equity risks. Without policy, AI use varies wildly by classroom. Some students get rich AI-integrated learning experiences. Others get blanket bans. The variation correlates with the same demographic factors that drive other educational inequities.
There are instructional risks. Without guidance, teachers either avoid AI entirely - depriving students of valuable learning - or integrate it haphazardly, without the pedagogical intentionality that makes technology integration effective.
A Path Forward
If your district is still in the early stages of AI policy development, here is my recommended path forward:
First, convene a representative task force. Include teachers from multiple grade levels and subjects, administrators, technology staff, a parent representative, and if possible, students. Diverse perspectives produce better policy.
Second, build shared understanding. Before anyone drafts policy language, the task force needs a common foundation of AI knowledge. This is where resources like iTeachAI Academy can help. When everyone at the table understands what AI can and cannot do, conversations are more productive.
Third, draft principles before rules. Start with the values that will guide specific decisions. Then develop context-specific guidelines that flow from those principles.
Fourth, pilot and revise. Implement the policy with a cohort of willing teachers, gather feedback, and refine before full rollout.
Fifth, commit to ongoing review. AI is not slowing down. Your policy process should not either.
Three years into the AI revolution in education, flying blind is no longer an option. Our students deserve clarity. Our teachers deserve guidance. And our communities deserve confidence that schools are navigating this moment with intention and care.
Janette Camacho, Ed.D., is the Founder and Chief Learning Architect at iTeachAI Academy (classes.iteachai.co). She is a Google for Education Certified Trainer and Coach, FETC 2024 and 2025 Featured Presenter, Adobe Creative Educator, Apple Teacher, and upcoming EdTech Digest State of EdTech 2026 Honoree with 28+ years of K-12 classroom experience.