Grafton AI task force outlines stoplight model, district rollout and training goals

The Grafton School Committee on June 10 received a detailed report from a cross‑functional AI task force that laid out a staged plan for using artificial intelligence in classrooms while prioritizing student data privacy and academic integrity. Kim Dufo, an elementary technology teacher, opened the presentation by describing a five‑step process that included landscape review, stakeholder input, district surveys, readiness evaluation and the drafting of guidelines and a roadmap.

Maren Mallaloy, the district technology coordinator, said the task force chose two primary domains to focus on after a January leadership summit: data readiness and security, and academic AI literacy. “We are focused on protecting student and staff data as we use AI tools across education,” Mallaloy said, and added the group intends to build “safe and compliant practices.”

The task force reported three stakeholder surveys (families, students, staff). Presenters said 54.9% of families identified student data privacy as a top priority and 44.1% of educators reported low familiarity with district expectations for AI. Presenters said student uses already include study guides, remediation for missed work and differentiated lesson planning, which reinforces the need for consistent district frameworks.

Festival Libren, the district’s AI integration and digital learning specialist, said draft guidelines emphasize AI literacy, digital citizenship, privacy and data protection, academic integrity, equity and bias awareness, and ethical decision making. Libren said the guidelines will be maintained as living documents and reviewed annually.

A central classroom policy described by task force members is a three‑tier “stoplight” model explained by Herdam Gotham. Under the model, assignments set to red prohibit AI use (work must be entirely original); yellow allows limited or structured AI use (for brainstorming or research); and green permits AI as a tool for efficiency and refinement while preserving student authorship. Gotham said teachers will communicate the stoplight level for each assignment to make expectations transparent.

Anna Fitzgerald, a middle‑school teacher who outlined phase‑two goals, said the district plans a districtwide expansion from August 2026 through April 2027. The task force’s implementation targets include 80% of staff completing training during the expansion, 90% positive stakeholder feedback and adoption of the stoplight model in all secondary classrooms that do not already use it.

On tools and data protection, presenters said the district is leaning on its Google for Education environment and related Google tools (including Gemini features available under the district license) because those services are provisioned under the district contract and do not use student information to train public models, according to the presenters. The task force also described securing a Commonwealth cybersecurity training grant and joining a data‑alliance arrangement to inventory district applications and require vendor data‑privacy agreements so the district can track where student data flows.

Committee members asked how teachers will be prepared to deliver consistent instruction and whether the district will direct students to specific tools. The task force said professional development will include role‑specific training, coaching and resources; a ‘‘battery’’ continuum will be part of staff training to help teachers judge how much a student is relying on AI versus demonstrating original work. Presenters emphasized transparency in source use and student vetting of research.

The committee praised the task force’s composition—students, arts teachers, educators and technologists—and stressed keeping a human in the loop. No formal vote was taken; the task force will continue refining its guidelines and communications plan ahead of the scheduled expansion.