Policy #5840 · Artificial Intelligence·First Read June 8, 2026
What It Is · What It Says · Where Reasonable People Disagree
The District's Proposed AI Policy, Explained
Clarence's school board is weighing its first formal policy on artificial intelligence—a framework that would let individual teachers decide whether students may use tools like ChatGPT, bar anyone from entering confidential information into them, and hold students and staff personally accountable for the accuracy of anything AI produces. Here is what the draft actually says, and the genuine questions it raises.
This policy is not yet in effect
Policy #5840 received only its first read at the June 8, 2026 meeting—it was presented for review, not adopted. The board is expected to vote on adoption at a later meeting, and the language below may change before then. This explainer describes the draft as presented and will be updated when (or if) the board adopts a final version.
In a Nutshell
- Who it covers: every student, employee, volunteer, and contractor in the district.
- The core rule for students: there is no district-wide ban or green light—each teacher decides whether GenAI may be used in their own classroom, and must explain their rules and discuss responsible use.
- The hard line on privacy: no sensitive, confidential, or proprietary information goes into any AI tool. The memorable test: if you couldn't release it under a public-records (FOIL) request, don't type it into AI.
- Accountability stays human: students answer for the accuracy of their work and staff answer for theirs—"AI is intended to support, not replace" a person's responsibility for their own work.
- It names the risks plainly: the draft itself flags AI's tendency to fabricate facts ("hallucinations"), to reflect racial and gender bias, and to raise unsettled copyright questions.
- Equity promise: the district says no student will be disadvantaged for being unable—or choosing not—to use AI.
What the Policy Says
Purpose & Scope
The draft frames AI—and especially generative AI (GenAI), the kind that produces text, images, and audio—as both an opportunity and a risk the district must "proactively address." It applies to everyone connected to the district and is written to sit alongside existing rules on data privacy, acceptable use, and student conduct rather than replace them. It explicitly anticipates revision as the technology changes.
Definitions
It distinguishes generative AI (systems that create new human-like content from large training sets) from traditional AI (fixed-rule systems that don't change behavior after deployment). The policy's substance is aimed squarely at the generative kind.
The Risks the District Names
Unusually for a policy document, the draft spends its longest section cataloging GenAI's weaknesses: hallucinations (confidently stated false information), bias (it explicitly notes that AI trained on human data can reproduce racial and gender stereotypes, ties this to the district's Diversity, Equity, and Inclusion goals, and warns of "response bias"—AI telling users what it predicts they want to hear), and copyright (unsettled legal questions about training data and ownership of AI output).
Data Privacy & Security
The firmest rule in the document: under no circumstances may sensitive, copyrighted, confidential, or proprietary information be entered into any AI tool. It anchors this to four laws—FERPA, NY Education Law §2-d, Labor Law §203-d, and State Technology Law §208—and offers a plain-language test anyone can apply: if it wouldn't be disclosed under a Freedom of Information Law request, it doesn't go into AI.
Student Use
The pivotal choice: rather than a single district rule, teachers decide whether their students may use GenAI for assignments. Teachers who allow it must clearly communicate expectations and discuss responsible use—risks, the student's accountability for accuracy, and proper citation. Misuse that conflicts with a teacher's instructions or the Code of Conduct can trigger investigation or discipline.
Staff Use & the Equity Pledge
Staff may use GenAI but remain "ultimately responsible for the accuracy and integrity of all work they produce." The district also commits to ensuring no student is disadvantaged by an inability—or a choice—not to use GenAI, an acknowledgment that access to these tools is uneven.
Training & Compliance
The district pledges professional learning and support for teachers (and possibly other staff), and ties AI use back to existing obligations—especially Education Law §2-d's requirements for any contract under which a vendor would receive student or staff data. The policy cross-references roughly a dozen related district policies on records, data security, and acceptable use.
Points of Debate
The Listening Post does not take positions. The following maps the strongest arguments on each side of the genuinely contestable choices in this draft—the questions a thoughtful board member, teacher, or parent might raise—so readers can weigh them. Purely procedural elements are not included here.
1. Letting each teacher decide whether students may use AI
The case for
Teachers know their subject and students best. A calculus teacher and a creative-writing teacher have legitimately opposite needs; a top-down rule would be wrong for one of them. Discretion lets policy track pedagogy, lets the district move now instead of waiting for a perfect universal rule, and treats teachers as professionals. It also keeps the policy durable as the tools change—the judgment lives with the educator, not in language that dates quickly.
The concern raised
The same freedom produces a patchwork: two seventh-graders down the hall can face opposite rules, and families can't predict what's allowed. Inconsistency is hard to explain to students and invites disputes over fairness in grading. It also quietly transfers a hard, contested policy judgment onto individual teachers without guaranteeing them the training or model rules to make it well—and "the teacher decides" can become a vacuum where the most confident (or least cautious) approach wins.
2. The DEI / bias language
The case for
Naming bias is substance, not decoration: AI demonstrably reproduces racial and gender stereotypes from its training data, and a district responsible for all students should say so and require staff to read outputs critically. Tying it to equity goals gives teachers explicit cover to reject a biased AI output and frames AI literacy as a fairness issue—protecting exactly the students a school is most accountable for.
The concern raised
Two directions of criticism exist. Some will argue embedding "DEI" framing in a technology policy is mission creep that imports a politically contested term into what should be a neutral operational document, and could date badly or invite controversy. Others will argue the opposite—that the language is aspirational but toothless: it describes bias and asks staff to "critically assess" outputs, but sets no standard, audit, or consequence, so it may not change what actually happens in classrooms.
3. The "FOIL test" for data privacy
The case for
It is the strongest, clearest line in the policy. Anchoring the rule to four named statutes makes it legally serious, and the "would you release it under FOIL?" test is a genuinely useful heuristic a busy teacher or clerk can apply in seconds without a lawyer. For a district holding sensitive records on children, a bright-line prohibition is the right default.
The concern raised
The FOIL test, while catchy, is an imperfect proxy—FOIL's exemptions are technical, and a well-meaning employee could misjudge what's disclosable. The rule also leans entirely on individual caution ("all users should exercise caution") without naming approved or prohibited tools, offering a vetted option, or describing enforcement—so the hardest part, knowing which platforms are safe, is left to each person to guess.
4. Holding students/staff accountable for AI's accuracy
The case for
This is the academic-integrity backbone. By making the human answerable for accuracy, citation, and integrity—"AI is intended to support, not replace"—the policy refuses to let "the AI said so" become an excuse, and teaches a skill students will need for life: verifying machine output. It preserves the value of the work without pretending AI doesn't exist.
The concern raised
Accountability without detection is uneven in practice: the policy presumes misuse can be identified, but AI text is notoriously hard to detect reliably, and AI-detection tools themselves produce false positives that can wrongly accuse honest students. Critics may also note the policy addresses honesty (cite your sources) far more than it addresses over-reliance—a student can use AI exactly as told and still learn less.
5. The equity pledge on unequal access
The case for
It is honest about a real divide—students don't have equal access to paid AI tools or home support—and commits the district not to penalize a child for that gap or for a principled choice to opt out. That protects lower-income students and respects families who object to AI on their own grounds.
The concern raised
It states a goal without a mechanism. "Will work with teachers to ensure no student is disadvantaged" names no funding, no district-provided tool, and no standard for what counts as a level playing field—so whether the promise means anything depends entirely on implementation the policy doesn't specify.
6. Copyright caution
The case for
Flagging that AI output may carry unsettled copyright and ownership questions is prudent and forward-looking; most district policies ignore this entirely. It signals to staff not to treat AI output as automatically free to use or publish.
The concern raised
It identifies the issue but offers no guidance—no rule on whether AI-generated material may be used in district publications, lessons, or branding, leaving the practical question open. As written it reads more as a disclaimer than a usable instruction.
Status & What's Next
Policy #5840 was presented as agenda item B12 at the June 8, 2026 meeting—a first read, which is informational. The policy committee chair singled it out as the one for board members to study before the next meeting, when adoption is expected. It was paired with a companion document, the district's Guidelines for the Appropriate Use of Artificial Intelligence, which sets out practical guidance separately from the policy itself. Eleven other policies were also first-read the same night; one—the Title I parent-engagement policy—was adopted early to meet a federal deadline.
How a Policy Becomes Final
Clarence policies typically get a first read at one meeting (presented, discussed, no vote), then are adopted at a later meeting—a built-in pause for board and public review. Policy #5840 is between those two steps. Watch the next BoE agenda's Board Development/Policy section for an adoption vote; the language could change before then.
Sources
[1] Clarence Central School District Board of Education, draft Policy #5840 (Artificial Intelligence), as presented for first read at the Regular Board Meeting of Monday, June 8, 2026 (agenda item B12). Full text in the official public board packet, clarence_BoE_20260608_agenda.pdf (4 pp. for this policy). This explainer paraphrases and summarizes the draft; it does not reproduce the policy text. The companion "Guidelines for the Appropriate Use of Artificial Intelligence" is a separate district document.
[2] Meeting context drawn from the June 8, 2026 BoE meeting brief (The Public Record).
Status note: This policy was not adopted as of June 8, 2026. Verify the final adopted text against the district's official policy manual once available.