East Penn rolls out AI policy, training and controlled staff access to Gemini

East Penn School District officials on Oct. 13 described districtwide work over the past year to adopt an AI policy, train staff and introduce controlled AI tools for classroom use while emphasizing teacher discretion and student privacy.

District leaders said the effort began after public release of generative AI in late 2022 and culminated with board adoption of an AI policy in September 2024. The policy covers academic integrity, evaluation of AI tools, ethical considerations and content verification, officials said. Technology Director Michael Billman and technology coach Dylan Peters outlined professional learning for staff, a student-facing video and a family technology resource page that lists approved tools.

Billman said the policy “addressed four key areas: academic integrity, evaluation of AI tools, ethical considerations, as well as content verification.” Peters described the training timeline, saying, “we have something that we need to get out in front of,” and recalled preparing teacher modules almost immediately after ChatGPT became public in November 2022.

The district said it produced a five- to six-minute video for secondary students that explains AI basics, safeguards against sharing personally identifiable information and includes a transparent note about how district staff used AI in the video’s planning. Students who watched the video could submit questions; Peters said the district received about 2,500 responses and invited students with substantive concerns to forums with technology staff. The district also provides an FAQ and an AI resources page for families.

Officials described a staged rollout for tools: staff and administrators have access to Google Gemini within the Google Workspace environment now, while student access is subject to additional evaluation. Billman described Gemini for Education as “safe and protected” and said it is not used to train Google’s models in the district environment. Staff also have access to vetted platforms such as NotebookLM and SchoolAI for moderated classroom activities, the presentation said.

On classroom use, the district adopted a traffic‑light model that leaves final discretion to classroom teachers: red = not allowed; yellow = allowed with teacher supervision; green = encouraged for specific learning objectives. Peters said teachers choose which students and lessons may use AI and that elementary use is teacher‑led; the district offered the secondary video to elementary homeroom teachers for guided viewing with grades 2–5 and teacher‑led demonstrations in K–1.

Board members asked about bias, student privacy, the potential for AI to analyze student assessment data and the ethical implications of AI use. Billman and Peters said the district evaluates tools for privacy and security and trains teachers not to include personally identifiable student data in general-purpose chatbots. Peters said district staff are teaching students to cross‑check AI output and to cite AI use when applicable.

Board members praised the district’s proactive approach and asked for continued reporting. Administrators said further training, teacher idea‑sharing and follow‑up presentations — including how AI could intersect with the district’s portrait of a graduate — are planned later this school year.

Ending

District leaders emphasized that their approach prioritizes teacher control, student privacy and iterative evaluation of tools as they consider opening Gemini and other systems to students in later phases.