The Question State Guidance Skipped

Paige Wyatt teaches high school English. Her administrators spent last year encouraging her to use AI to write lesson plans faster. Her students, in the meantime, used AI to write their papers so often that she started paying out of her own pocket for detection tools.¹ That contrast, described in a Chalkbeat review published on September 2, is the quiet scandal of AI in schools right now. Guidance keeps arriving. It keeps skipping the question that matters most.¹

The Chalkbeat review looked across state after state. Most have now issued documents about AI in schools. Most dedicate pages to AI literacy, to sample prompts, to ways teachers might use chatbots to draft rubrics or newsletters. Most give the question of academic integrity a paragraph or two, sometimes just a bullet, before moving on.¹ What they do not do is give a teacher any concrete way to know whether the essay in front of her was written by the student whose name is at the top.

That silence sits on top of a number no policy can ignore. In August, Common Sense Media published its second census of teen AI use. Seventy percent of teens now use AI for schoolwork.² Among those who do, close to two thirds say they are asking the model for the answer, either pasting it in as their own, rewriting it to sound like themselves, or lightly editing it.² Only three in ten teens report that a teacher has ever discussed with them how to use AI honestly.² The story from Wyatt's classroom, in other words, is not one classroom. It is the median classroom.

Why the Guidance Goes Quiet

It is tempting to read the state guidance as cowardice. That is probably unfair. The reason academic integrity gets one page in most of these documents is that no one, honestly, knows what to say. Detection tools are unreliable.¹ Bans are unenforceable. Take-home essays as an assessment form are quietly collapsing. Officials asked to write policy are being asked to solve a problem that policy alone cannot solve. So they write around it.

But writing around a problem does not make it smaller. It just moves it into the room where the teacher is standing.

Where the Future Is Actually Being Decided

The future of education is being decided in that room, not in the ministerial statements or the vendor keynotes. It is being decided in the pause between when a student opens a blank document and when a finished essay appears in the teacher's inbox. Everything the profession is going to have to relearn about assessment lives inside that pause. Whether it happened at all. What happened during it. Whether the person who typed the last sentence was the same person who wrote the first one.

The old system did not need to watch that pause because the pause was largely invisible and largely irrelevant. A grade was a summary of a product. The product was the point. That premise has quietly stopped being true. When a machine can produce a serviceable essay in fifteen seconds, the product tells you almost nothing about the person whose name is on it. What the product used to prove, the process now has to.

A Different Question

A school that takes this seriously will stop asking, "Did a student use AI?" and start asking, "What did their thinking actually look like?" Those are very different questions. The first has no reliable answer and turns teachers into forensic investigators. The second has a clear answer, provided you can see the work as it was made. The drafts. The revisions. The pauses. The moment a student asked for help and the moment they closed the tutor and finished the sentence themselves. Not surveillance. A record. The kind of record a serious teacher would already keep for a student they were mentoring, if she had time.

Koan is built on the premise that the record is now the point. Not the finished essay, which any tool can produce, but the visible shape of a student becoming capable of producing it. When a school can see that shape, most of the anxiety around cheating quietly dissolves, because the honest work looks different from the copied work, and everyone in the room can tell.

The state guidance documents will keep arriving. Some will get braver. Most will keep skipping the question. The teachers cannot skip it. Sixty seconds after the bell rings, they are already inside it.

So maybe the useful question, the one no state playbook has yet put on paper, is this. If you cannot see how a student thought, what exactly are you grading?