When GenAI arrived in schools, Fairhaven's superintendent came to Kat Ware with a simple question: what do we do?

Kat Ware is Fairhaven Public Schools' EdTech Coordinator, a role the superintendent created for her after an interview process that made it clear that the open AP position didn’t feel like the right match. "She always finds the right people to fit the right roles," says Jennifer Perrotti, Fairhaven's Innovation Coordinator and Kat's close collaborator. "She has a knack for that." The superintendent's instinct was to get ahead of things she knew were coming down the pike. Kat's instinct was to build something grounded, evidence-based, concrete, and relevant enough to be used across district stakeholders.

Fairhaven currently has a comprehensive AI guidance system that was built by a small team in a single year.

First Things First

Before Fairhaven built anything, they figured out where they were.

The first step was a district-wide staff survey, designed to understand what teachers were already using, what they needed, and what they were worried about. "Any decision made at the district level has to be backed by something," Kat says. The survey gave them data to inform the work.

The second step was naming the absolutes. Before drafting a line of guidance, the team identified the non-negotiables. Student safety was first. And the framing that emerged from early conversations became the backbone of everything that followed: Human, AI, Human. Whatever the tool, whatever the task, a human being initiates and a human being closes.

From January through March, the team worked through iterative drafts, running content back through the AI team, incorporating feedback, and comparing their language against DESE guidance, other district policies, and relevant Massachusetts state law. Kat also drew on her dissertation research, which grew to include eight or nine books and nearly 500 articles on AI in education. 

What They Built

Fairhaven's AI Resource Hub is designed to serve every member of the school community, from a middle school science teacher wondering what tools are safe for students, to a parent who wants clarity about when a student can use AI to help with homework. 

The website opens with the district's guiding principles and Fairhaven's Human-AI-Human approach, then branches into guidance for students, teachers, and administrators. It includes a curated list of tools vetted against the Student Data Privacy Consortium agreements, with clear distinctions between what teachers can use and what students can use. A family and community section is planned for launch once the policy is approved.

An AI Assistant lives at the top of the site. Kat built it as a Gem, trained exclusively on Fairhaven's own documents. It does not pull from the internet or hallucinate. If it doesn't have an answer, it routes the user to the appropriate human: a teacher, a principal, or the technology coordinator, depending on the question. A parent, teacher, or student with a question can get a clear answer based on Fairhaven's actual policy without emailing anyone.

The policy itself is also a notable artifact. Structured around a five-level AI Acceptable Use Scale, it gives teachers a clear framework for communicating expectations at the assignment level: from Level 0, where no AI assistance is permitted, through Level 4, where AI is intentionally integrated and students demonstrate original reasoning alongside it (which was adapted from Ryan Robidoux’s Idea Ownership Spectrum). The policy also establishes a cross-departmental AI Governance Committee, a phased implementation framework moving from exploration through piloting, scaling, and sustaining, and unusually specific prohibited uses for both staff and students. Among the most forward-thinking: an explicit prohibition on uploading or entering any visual likeness of a student into an AI tool, including photos, screenshots, and video stills, which is grounded in FERPA, COPPA, and Massachusetts student records law. 

Kat’s AI Policy Literature Review is the research backbone of the system. Built from her dissertation work, it provides substantive, sourced answers to school committee members, community members, and anyone else who asks for the research behind the district's decisions. 

All content is designated as “working documents”. The team plans to review the AI policy monthly even after it is officially published, so that guidance evolves alongside the technology. Any updates to the policy will flow into the supplemental content resources. 

Conditions for Success

The decision to create a role specifically for Kat, rather than place her in an existing position, was a strategic investment. A forward-thinking superintendent comfortable with disruptors created the conditions for this work before the work began.

Jen identified Fairhaven’s team composition as one of the most important conditions behind their progress, and named important roles specifically: upper administration, Kat sitting between the tech and curriculum worlds, Jen representing elementary voice, the technology director, the curriculum coordinator (who leads PD), and the assistant superintendent (to help move policy and guidelines forward with the Policy Committee).

Kat's background was also unusual in ways that mattered for this work. She came to the EdTech Coordinator role with a principal's license, a specialization in EdTech leadership, years of classroom teaching experience from kindergarten through 12th grade, active involvement in an early ChatGPT programmer's network, and a doctorate (in progress) focused on AI policy in education. She understood the technology from the inside, the classroom from experience, and the policy landscape from research. That combination is rare, and it shaped what Fairhaven was able to build. 

Including elementary perspectives was something both Kat and Jen flagged as a differentiator. Many districts treat AI as a middle and high school concern and Fairhaven pushed back on that instinct. Elementary students need to start building AI literacy the same way they need digital literacy, and developing guidance without an elementary perspective doesn't support the full K-12 planning that districts need to consider. 

The PEA2K cohort served as a forcing function for the team's progress. “PEA2K has helped us move AI from conversation to action," Jen says. The structure of the cohort, the monthly sessions, the coaching, the deadlines, the resources, created conditions for progress that the team might not have created on their own. "It kind of grabbed us and said, okay, you need to get this done."

Recommendations 

Build a Foundation. The instinct when facing pressure to act on AI is to move quickly. Fairhaven's experience suggests the opposite: invest time upfront in building a team, name your values, survey your community, and agree on your non-negotiables. The work will move faster once that foundation is in place. 

Design for the whole community. The Fairhaven team designed for K-12 teachers, students, administrators, families, and school committee members from the start, building community trust and common understanding. 

Treat everything as a working document. AI is changing too fast for any guidance or policy to be permanent. Building in a regular review process is the most honest response to the pace of the technology.

As Jen puts it: "No one has all the answers, but transparency is the best first step."

This case study was developed through an interview with Kat Ware, EdTech Coordinator, and Jennifer Perrotti, Innovation Coordinator, at Fairhaven Public Schools, in June 2026. It was prepared by Cathy Sanford, Director of Communications at Throughline Learning as part of PEA2K (Partnership for Educators Advancing AI Knowledge) Cohort 1 documentation.