🧠 AI Literacy Lab HomeDeckPortal

Email 4 — Brief to Leadership / Board

Send after the session. This is the ask.


Subject: AI training outcomes — and 4 decisions that need you


[SUPERINTENDENT / CABINET / BOARD],

Summary of the AI professional learning day on [DATE] and the items that require
district-level action.

What staff can now do

[N] staff completed the session. The classroom framework is built on three pillars —
Clarity, Capacity, Community — and every element maps to the NJDOE's published
principles for K-12 AI use, including its own principle to *"prioritize process over final
products in assessment."* This is not a local invention; it is state guidance operationalized.

Teachers left with a labeled unit, a redesigned assessment, and a conversation protocol.
Follow-through is structured at 30/60/90 days.

What teachers cannot fix, and need you to

The session was explicit with staff that the following are not teacher-level problems.
Being honest about that was essential to credibility, and it now needs to be honored.

#DecisionOwnerWhy it matters
1Enable pacing gates in the credit-recovery platform — units unlock sequentially with a minimum-days ruleTechnology + CurriculumThis single configuration stops most of the "six months in six hours" pattern. Highest return of anything on this list.
2Extension management on student devices — move to an allowlistTechnologyUnmanaged browser extensions are the largest technical bypass vector. Also a CIPA-adjacent question.
3Publish an approved AI tool list with signed data agreementsTechnology + CounselRequired under P.L. 2019, c.494. Teachers cannot comply with a list that doesn't exist. Right now staff are making individual decisions with no guidance, which is the actual risk.
4Review credit-recovery policy: is credit earned by completing modules or by demonstrating mastery?Curriculum + BoardIf it's the former, the incentive to speedrun is structural and no teacher can out-teach it. This is the root cause.

Two policy provisions we recommend adopting

1. No disciplinary action based solely on an AI-detector score.
Detector outputs are probabilistic, trivially defeated by freely available "humanizer"
tools, and produce false positives at elevated rates for multilingual writers. In a district
with Irvington's linguistic diversity, a detector-driven discipline process would not
distribute its errors evenly — it would concentrate them in a protected class. That is a
Title VI exposure, not merely a technical shortcoming. A written provision protects students
and protects the district.

2. A documented appeal path for any student contesting an AI-related integrity finding,
plus tracking of AI-related referrals by subgroup. If disparities emerge, we want to
have found them ourselves.

Curricular note — this is already required

P.L. 2022, c.138 (S588) made New Jersey the first state to require K-12 information
literacy instruction, defined as the ability to locate, evaluate, and effectively use
information. Teaching students to evaluate AI output is that requirement, applied to the
information source they now use most. It maps to NJSLS 9.4.x.IML, 9.4.x.DC, and
8.1.x.IC.

This means AI literacy is not an unfunded add-on competing for time. It is an existing
obligation, and framing it that way resolves the "where does this fit" objection before it
is raised. The statute also names school library media specialists as curriculum
partners — they should be formally included.

Possible funding

NJDOE has operated an Artificial Intelligence Innovation in Education Grant supporting
both Teaching with AI and Teaching about AI, including AI literacy curricula and teacher
professional development. Worth evaluating for the 60/90-day work.

What I need from you

  1. A decision on items 1–4 by [DATE]
  2. Counsel's review of the two recommended policy provisions
  3. Ten minutes at the [DATE] Board meeting to present the framework

Full documentation attached. The compliance checklist in the standards crosswalk is written
for counsel and the data-privacy officer.

— [NAME]
[TITLE]


*Legal citations in the attached materials are summarized from public sources for
professional learning purposes and should be confirmed by counsel before policy adoption.*