The Line The Adult Forgot To Delete

This week the Charleston City Paper sent three campaign questions by email to Leah Whatley, a Charleston County School District trustee running for re-election in District 7. One of her answers came back beginning with a sentence that was not an answer. It read, "Recommended response, assuming you are not seeking one." Then a formal denial about seeking an endorsement from Moms for Liberty. When pressed on the obvious use of AI, Whatley told the paper, "AI is an amazing tool when used correctly."¹

The same week, her district rolled out a districtwide AI policy for the new school year. It is called the SEE framework. Safe, Ethical, Effective.² Every teacher was trained over the summer.³ Every student, this fall, will be introduced to it. And the new code of conduct singles out one behavior in particular. Using generative AI to develop content and presenting it as if it is original to the user without proper disclosure.²

The staff of the paper noticed the sequence and wondered aloud what would happen to a Charleston student who submitted a term paper the same way.¹ That is the useful question. Not the political one about the trustee. Not even the ironic one about a board member breaking the rule her own district was rolling out on the same Monday. The useful question is a smaller one. What is the rule actually about.

The Missing Layer

Read the trustee's response one more time. "Recommended response, assuming you are not seeking one." That is not the answer. That is the scaffolding underneath the answer. It is the AI's suggestion, sent to a human, meant to be considered, revised, or discarded. The mistake in the email was not that a trustee consulted a tool. It was that the layer between the tool and the reader was missing. She sent the scaffolding as if it were the building.

The disclosure rule Charleston wrote is a rule about that layer. Not about whether AI touched the work. About whether the human touched it back. A term paper written with AI, revised by a student, and disclosed honestly is one thing. A term paper generated by AI, copied without a look, and turned in as original is another. The word Charleston chose is "disclosure." The word underneath the word is "look." Did the student look at what the machine gave her before she put her name on it. Did she change any of it. Did she leave any of herself in it.

What Adults Model First

The awkwardness of the Charleston story is that a school district's first AI policy almost always instructs children before it instructs the adults around them, and children read the room before they read the rulebook. When a trustee sends a scaffolded response with the scaffolding still attached, the ninth grader whose English class starts next week has already been given the demonstration. That the rule is easy to break. That the layer is easy to forget. That the phrase "AI is an amazing tool when used correctly" can be used as cover for not having used it correctly.

There is no way for a policy to reach into every adult email in a district. There is a way for a policy to name what a real answer looks like. The name Charleston is using is a good one. Safe. Ethical. Effective. Effective is the word that carries the most weight in this particular story. Effective means the user looked at what she got. It means she used the tool, rather than being used by it.⁴

The Record That Would Have Helped

At Koan we think about the record a student leaves as she works. The pauses. The revisions. The moments the sentence changed. Not because that record catches cheating, though it can. Because that record is the layer itself. A teacher who can read the layer already knows the difference between a paper the student sat with and a paper the student passed through her. The trustee's answer, viewed in that light, was not scandalous because it was AI. It was thin because it was one layer. A note from the machine, forwarded with no note from the human.

The lesson for a school this month is not about detection tools. It is about the habit of leaving the human layer visible. If a rule about disclosure is a rule about a look, then what a school is teaching is how to look at something before signing your name to it. What it is trying to make visible is not the tool. It is the person on the other side of the tool. The student. The teacher. The trustee. The adult whose email came out sounding like a suggestion because a suggestion was all it was.

If a student and a trustee had submitted the same email this week, would the school's policy have judged them by the same standard, or would it have judged them by two?