Lifetime Citizen Portal Access — AI Briefings, Alerts & Unlimited Follows
Policy committee advances draft on generative AI, urges clear copyright and recording rules
Loading...
Summary
The Upper Dublin School District policy committee moved draft policy 815.1 (use of generative AI in education) to legislative first reading after debating copyright safeguards, blocking of non‑endorsed tools, audio/video recording limits, and user responsibility for verifying AI outputs.
The Upper Dublin School District policy committee advanced draft policy 815.1, governing the use of generative artificial intelligence in classrooms, to the legislative agenda for first reading.
Ms. Kitten introduced the item as part of a broader renumbering effort to align district policies with Pennsylvania School Boards Association (PSBA) guidance. Ms. Evans, who led the policy presentation, said the draft is intended to provide "guardrails for us" and acknowledged the policy will likely need frequent review as tools evolve. "We view this policy really as guardrails for us, and we recognize that it will most likely maybe this is 1 that needs to come back every year to have to be reviewed just because so much is changing in it," Evans said.
Committee members pressed for clearer operational guidance. Mr. Lee urged more directive copyright language so students and staff "should not copy and paste, upload, or manually input any copyrighted material into a publicly accessible AI tool," explaining that material pasted into public-facing models can become part of the tool's dataset. Evans acknowledged the concern and said the district is drafting an administrative regulation and student‑facing handbook guidance to address implementation details.
The draft adds a new section (3.6) on audio and video recording. Evans explained the intent is to prohibit capturing or summarizing meetings or individuals via automated recording/summary tools without consent. She told the committee that while a recording device is hardware, many platforms include AI components that can transcribe or summarize content, and the policy seeks to prevent unauthorized capture and distribution.
On technical controls, Evans said the district cannot block specific tools at the network level but "through our Securely software, we can block by category and then unblock the tools that we are endorsing," enabling the district to permit vetted, licensed tools while limiting public‑facing systems.
Members also discussed verification and accuracy. Ms. Lowe asked for language placing responsibility on the user to ensure the validity of AI output; several members agreed to add a requirement that users apply "proper oversight and evaluation of generated information for validity" and cross‑reference trusted resources.
The committee voted to move the draft policy to legislative for first reading to allow broader board review; the committee reasoned the policy's guardrail approach, combined with an AR and periodic review, would balance safety and instructional innovation.
Next steps: the administration will draft the administrative regulation (AR) and student/staff guidance referenced during the discussion for the board packet at legislative first reading.
