All activities Professional learning
Leadership teams start with what they value, sort draft clauses, and write a one-page generative AI guidance statement that people can actually follow.
Many campuses have AI guidance that is either a ban nobody enforces or a list of tools nobody reads. This session builds guidance the other way around: teams choose the values that matter most for learning on their campus, sort a deck of draft clauses to find which ones are clear, fair, and enforceable, and then draft a one-page statement covering acceptable use, disclosure, privacy, and equity. Along the way they critique an AI-generated draft, which models the same judgment the guidance asks of students and staff.
Hook
Read three quick scenarios aloud: (1) a student uses AI to brainstorm essay topics and doesn't mention it; (2) a teacher pastes a student's IEP goals into a chatbot to draft accommodations; (3) a family can't afford the premium version of an AI tool a teacher recommended. For each, ask: "Does our current guidance tell this person what to do?" People answer with thumbs up, sideways, or down.
Facilitator noteMost rooms go sideways or down on all three. That tells you the need: guidance that answers real situations, not guidance that lists tools.
Explore
Each team spreads out the twelve values cards. In silence, each person places a sticky dot on the three values that matter most for learning on their campus. Then the team talks: keep four, set aside the rest. For each kept value, the team writes one sentence on chart paper: "On our campus, this means…"
Facilitator noteListen for values that pull in different directions, like Curiosity and Safety. Don't resolve the tension. Good guidance names it.
Practice
Teams open the Handout B envelope and sort sixteen draft clauses onto the mat by the area each belongs to: Acceptable Use, Disclosure, Privacy, or Equity and Access. Then, within each area, they flip over (or mark with an X) any clause that is vague, unenforceable, or quietly unfair, and rewrite one of them on the back. Compare with the facilitator key.
Facilitator noteWatch for teams that keep "AI detection software will be used to identify cheating." Ask: what happens to a student falsely flagged? Detection tools are not reliable enough to be the only evidence.
Model
Project the AI-generated guidance statement you prepared. Ask teams: "Which of your four values does this draft honor? Which does it ignore? What would it tell the three people from our opening scenarios?" Mark up the projected draft together, highlighting generic language and missing pieces.
Facilitator noteUnplugged or no approved tool: describe a typical AI draft aloud. It is usually fluent and well organized, bans "using AI to complete assignments" without defining that, says nothing about staff use or student data, and never mentions cost or access. The lesson is the same: a polished draft is a starting point, not a decision.
Create
Teams use Handout C to write a one-page guidance statement. Every clause must trace back to one of their four values, and the draft must answer the three opening scenarios. Teams also fill in how students, staff, and families will learn about the guidance and when it will be reviewed. Post drafts on the wall when finished.
Facilitator noteEncourage short, plain-language clauses. A good test: could a seventh grader and a parent both explain it back to you?
Debrief
Teams rotate to another team's draft and test it against a new scenario you read aloud: "A student with dyslexia uses AI text-to-speech and an AI outline tool as supports. A classmate says that's cheating." Visitors write one sticky note: what the draft says to do, or what it's missing. Teams return, read their notes, and name one revision.
Facilitator noteThis scenario tests whether the equity and acceptable-use clauses talk to each other. Guidance that treats all AI use as a single thing fails here.
Transfer
Each team names the next three groups who must review the draft before it is adopted, including at least one student group and one family group, and a date. Each person writes one commitment on the back of Handout C.
Facilitator noteRemind teams to check their draft against existing board policy and district data privacy agreements before sharing widely.
Everything except the AI-draft critique runs on paper, and that step works by describing a typical AI draft aloud (see the Model step note) or by printing an AI draft you generated in advance. The values sort, clause sort, gallery walk, and sticky-note stress test are all physical and are the heart of the session.
Run the values sort with a digital sticky-note board and dot voting, and the clause sort as drag-and-drop cards in a shared slide. Teams draft in a shared document with comment access for the gallery walk. Only the facilitator uses the AI chatbot, projected, so no participant needs an account. In a virtual session, use breakout rooms for teams and have visiting teams leave comments instead of sticky notes.
Does it need a screen? Critiquing a live AI-generated draft makes the point that fluent is not the same as fit for purpose, and it models the verify-and-decide habit the guidance asks of everyone. Shared drafting documents also leave a revision history that shows how feedback changed the statement.
What you should be able to see or collect if it worked.
Teams draft campus AI guidance on disclosure and privacy, the expectations students meet in the digital citizenship ethics-and-laws and privacy-and-security substrands.
Share the draft with one student group and one family group within two weeks and ask them to explain it back in their own words. Revise anything they misread, then bring the draft to your leadership team with the feedback attached.
