Teaching Healthy AI Use
The difference between AI that builds a student’s capability and AI that replaces it is a teachable distinction — and it is not the one most AI training covers.
Why this sequence exists.
The five Healthy AI Use guides each already have an educator page — misconceptions, discussion prompts, and classroom activities, built and ready. What they have never had is an order, a time budget, and a way to tell whether any of it worked. That is what this page adds.
Three findings set the shape of it:
69% vs. 47%
Teachers report no formal guidance for one-on-one instruction and tutoring at 69%, against 47% for producing worksheets. Guidance is most available where AI saves a teacher time and least available where it touches a student — which is the only place it changes what a student learns.
Design removed the harm
In the randomized trial where unrestricted AI raised practice scores 48% and lowered the later unassisted exam 17%, the version that eliminated the harm was one whose safeguards were written by two of the school's own math teachers. The variable that decided the outcome was instructional design, not the model.
A good tool, unused
In a two-year randomized trial across 18 middle schools, an AI tutor deliberately configured to coach produced gains resembling the same platform without AI. 96% of students tried it; the median student messaged it in only 17% of the sessions where they made a mistake. The tool was built correctly. Teaching anyone to use it that way was a separate job, and nobody did it.
The full evidence, with citations and where federal and state law currently stands, is in the issue brief: Preparing Educators to Teach AI Use.
The six-session arc.
Sessions 1–3, then a four-week practice period, then sessions 4–6. Each of the first five sequences an educator page that already exists, and each one ends in a decision a teacher can make about a specific task. The sixth is where those five decisions get made at once, for a real task, and written into the short paragraph students will actually read — the AI Use Statement.
Extraction vs. Scaffolding
~60 minThe usage-mode distinction, and the whole foundation of the sequence. Everything after this assumes the room has stopped arguing about exposure and started asking about mode.
- Activity
- The twin-task demonstration — half the room works a problem set with AI as an answer machine, half with AI as a hint-giver, then everyone does a related problem unaided. It re-runs the study the guide is built on, at classroom scale.
- Misconception
- That the amount of AI is what matters.
What You Hand Off
~60 minDeliberate versus habitual offloading. The useful move here is the keep-sharp list: the capabilities a teacher wants their students to still have in five years, named out loud.
- Activity
- The Offload Audit, run on the participants' own work first and their students' second — which capacities are being handed over, and were any of those hand-offs a decision?
- Misconception
- That offloading itself is the problem.
The Perception Gap
~60 minWhy students — and teachers — cannot feel it happening, and why that turns measurement into a design requirement. This is the session that motivates the pre/post.
- Activity
- The Calibration Check as a group read — estimate before the result, then compare. The teachable beat is how badly the estimate misses over three minutes, and what that implies for judging a skill over a semester.
- Misconception
- That a teacher, or a student, can feel skill slipping if they pay attention.
Four-week practice period · one structured check-in
This is the differentiator, and it is the part most likely to be cut. Between sessions 3 and 4, participants teach for four weeks using what they have — one deliberate change to how they set up AI use in their own classroom — and come back to a single structured check-in at the halfway mark: what you changed, what happened, what you could not tell.
NCDPI’s own generative-AI guidance recommends guided practice followed by a practice period of at least four to six weeks. The professional-development research that has actually been measured against student outcomes associates effect with sustained duration, coaching, collaboration, and modeled practice — while the format that dominates real delivery is the one-off session. In one 2025 survey, half of teachers had received at least one AI session and a third of those attended only a single one.
Said plainly: a single asynchronous pass is not what the evidence supports. If the practice period is removed, this becomes a different program, and it should be called one.
Effort Is the Active Ingredient
~60 minDesirable difficulties, and why fluency misleads. Run after the practice period, because teachers arrive with four weeks of their own examples of friction they removed or protected.
- Activity
- The Mastery Map, applied to a unit the participant is about to teach — where is the productive struggle, and where would AI quietly remove it?
- Misconception
- That struggle is either always good or always something to relieve.
Boundaries
~60 minWhere AI use stops being about learning. This is also the session that gives teachers language for the conversation the national data says almost none of them have been prepared for.
- Built on
- Boundaries — educator guide
- Activity
- The Boundary Check, plus the displacement question applied to a class: is AI adding to what students do with each other, or replacing it?
- Misconception
- That heavy use is the warning sign.
The AI Use Statement
60 min · working sessionA working session. The five sessions each produce one decision about a task; this is where those decisions get made once, together, for a real task, and written in a form students actually read. It is the only session that produces an artifact — and it is deliberately small, because a statement a teacher can write again in ten minutes for the next task is worth more than one redesigned assignment.
- Built on
- Built for this pathway
- Activity
- Each participant picks one task they are actually assigning in the next month and writes the five-line AI Use Statement that will go on it — the short paragraph telling students what to use AI for, what stays theirs, and where AI stops. Each line comes from one of the five sessions. A partner then plays the student and tries to satisfy the statement while doing as little of the protected thinking as possible; wherever that succeeds, the statement gets tightened.
- Misconception
- That the sequence is about deciding how you feel about AI.
The five lines — one per earlier session
- Mode
- “On this task, use AI to ___. Do ___ yourself.”Session 1 · Extraction vs. Scaffolding
- Protected
- “When this is over you should be able to ___ without AI.”Session 2 · What You Hand Off
- Signal
- “Here is how I will see whether that happened: ___”Session 3 · The Perception Gap
- Effort
- “This part is supposed to be hard, and it stays hard: ___”Session 4 · Effort Is the Active Ingredient
- Stop
- “AI use on this task stops at ___.”Session 5 · Boundaries
Measuring it.
Pre / post
CPAI Proficiency Assessment — Administer it before session 1 and again after session 6. It is the only instrument here that produces a comparable number at two points in time.
Formative, per session
Each guide's interactive doubles as a low-stakes check for understanding — Mode Check, Offload Audit, Calibration Check, Mastery Map, Boundary Check. No accounts, nothing stored; run them as a group read and use the disagreements as the discussion.
The artifact
Session 6 produces one AI Use Statement per participant, written for a task they are about to teach. It is the sequence's only deliverable, and the one a department head can actually read four weeks later to see whether anything changed.
The honest caveat, stated up front
A 2026 review that screened more than 800 papers and identified 20 high-quality causal studies of AI in K–12 found none evaluating AI-focused teacher professional development. There is no rigorous evaluation of a program like this one to point at — not ours, not anyone’s.
So anyone running this sequence becomes part of how that evidence finally gets made. If you run it and measure it, we want the data, and we will publish what it shows.
Facilitator packet.
Session-by-session facilitator notes, every misconception and discussion prompt from the five educator pages in one place, the practice-period check-in protocol, and a one-page handout of the four findings. Print it or save it as a PDF. It lives at a URL, so it never goes stale in someone’s drive.
Open the facilitator packet →Where this leads
Running this with a whole faculty is a different job.
CPAI delivers this pathway directly to schools and districts, prepares instructional coaches to run it themselves, and adapts the same material for students. The policy evidence behind it sits in the Education & Schools briefs.