The Cherry Hill Board of Education adopted Policy 2365 on Oct. 13, 2026, establishing a five-level framework for generative AI use across the district. The policy applies to all 10,919 students in pre-K through grade 12 and covers district-owned and personal devices on and off school premises [1] [4].
The board considered but dropped a parental opt-out that would have let parents require no AI for their child, according to reporting by the Philadelphia Inquirer and the Cherry Hill Courier [3] [4]. Officials said the SchoolAI closed platform addresses safety, and equity concerns about disadvantaging students barred from the technology weighed against the opt-out. The version of Policy 2365 currently posted on the Cherry Hill Public Schools website is marked "DRAFT" and dated March 24, 2026; that draft still includes a provision allowing parents to submit a formal letter to the principal requesting Level 0 for all assignments [1]. We are still verifying whether the opt-out provision was removed in the final adopted policy or retained in modified form.
Five-level framework
Policy 2365 requires each assignment to carry a designated AI Use Level from 0 to 4. Level 0 allows no AI. Level 1 restricts AI to ideation only, with no AI content in final submissions. Level 2 permits AI-assisted editing of the student's own work. Level 3 allows AI task completion but requires human evaluation and adaptation. Level 4 allows full AI integration, including drafting, with disclosure, prompt logs, and evidence of critical thinking. If no level is stated, the default is Level 0 [1] [2] [7].
Grade-band ceilings
- Pre-K and kindergarten: no independent student AI; teacher-facilitated demonstrations only.
- Grades 1 through 2: teacher-directed whole-class interactions; students do not operate AI independently.
- Grades 3 through 5: supervised individual or paired interactions for defined tasks; grade 4 includes brainstorming and limited editing; grade 5 expands ideation and formal citation begins.
- Grades 6 through 8: full range of AI-assisted workflows with increasing independence; full citation and disclosure required; high-stakes assessments default to no AI.
- Grades 9 through 12: full spectrum of AI use as staff designate; high-stakes and summative assessments default to no AI; AP students follow the more restrictive of district or external standards.
Citation and disclosure requirements
Disclosure rules escalate by grade. Grades 2 and below require a verbal or simple written statement. Grades 3 through 5 require a written indication of AI involvement and how it was prompted. Grades 6 through 7 require naming the platform, describing the prompt, and including the URL. Grade 8 adds the date. Grades 9 through 12 require MLA 9th Edition citation format for AI interactions [2].
Prohibited uses
- Submitting AI-generated work as one's own at any level not authorized.
- Exceeding the designated AI Use Level.
- Failing to disclose AI use when required (Levels 1 through 4).
- Using AI to generate or disseminate false or misleading information.
- Creating, possessing, or distributing AI depictions of real people to harass or bully (deepfakes).
- Using AI with confidential student or staff personal information.
Opt-out provision considered
The board reviewed a parental opt-out that would have allowed parents to submit a formal letter to the principal requesting Level 0 for all assignments. It drew criticism over equity concerns, with some arguing that students barred from AI would be at a disadvantage. Superintendent Kwame Morton said the district concluded the opt-out was unnecessary because SchoolAI's closed monitoring system provides safety [3] [4]. Director of Curriculum and Instruction Allison Staffin said pre-K and kindergarten students are not permitted to use AI independently [3]. As noted above, the draft policy document posted on the district website retains the opt-out language, and the final adopted version has not been independently reviewed by Formative Spaces [1].
Haddonfield's elementary ban
In August 2026, the Haddonfield School District, which enrolls about 2,700 students in pre-K through 12th grade, adopted a three-tier policy that bans generative AI outright for elementary students in pre-K through grade 5 [3] [6]. AI literacy lessons begin in first grade during library time [6]. Middle school students in grades 6 through 8 use AI only through Gemini Notebook under supervision [6]. High school students use AI with teacher guidance on approved platforms [6]. Assistant Superintendent Gino Priolo said the policy bans younger students from using AI so that they learn foundational and critical learning skills before being introduced to the technology, the Inquirer reported [3].
Unlike Cherry Hill's graduated approach that permits teacher-facilitated AI use starting in pre-K and supervised use by elementary grades, Haddonfield imposes an absolute prohibition through grade 5. Both districts default high-stakes assessments to no AI.