How Teacher Visibility Changes the AI Integrity Conversation

Prohibition without visibility is a standard without infrastructure

THE PROBLEM WITH CURRENT POLICIES

Most middle school AI policies prohibit AI-assisted work. The standard is correct. The enforcement mechanism is not. Detection tools are unreliable and adversarial. In a standard homework-to-submission workflow, the teacher sees the product, not the process. AI-generated text that mirrors a student’s voice cannot be distinguished from the student’s own work at the product level.

WHAT VISIBILITY CHANGES

When a teacher can see the drafts, the feedback, and the revisions, not just the final submission, the question shifts from “did they cheat” to “what did the work actually produce.” Process visibility changes the cost structure for students: submitting AI-generated work requires fabricating a plausible revision history, which requires more effort than doing the work.

THE ARCHITECTURE THAT PRODUCES VISIBILITY

School-governed: the teacher controls what tool is used, under what configuration, and in what context.

Class-linked: teacher access to student work is automatic and does not depend on student disclosure.

Assignment-bounded: student work stays within the scope of the teacher’s task, not extended to open-ended AI use.

Draft history preserved: the teacher can see how the final product developed and what changed across revision cycles.

WHAT THIS LOOKS LIKE IN GUIDED SCHOLAR

Guided Scholar operates on a school-governed architecture. Students access it through class-linked accounts against teacher-configured assignments. The system delivers structured, rubric-aligned feedback, not open-ended AI assistance. The teacher dashboard shows every submission, every feedback cycle, and how the draft changed in response. The question shifts: not “did they use AI” –which detection cannot answer—but “what did the student do with the feedback, and did their writing improve.”