AI visibility

    AI visibility, not AI detection.

    AI visibility is the practice of recording how a student worked with AI — every prompt, every response, every revision — so a teacher can assess the process rather than estimate whether the finished text was machine-written.

    A detector puts you in an argument with your student. A record puts you in a conversation.

    What can a teacher use instead of an AI detector?

    A record of how the work was done. Three things you need in order to mark it, none of which a probability about the finished text can give you.

    What the student actually did

    Which sentence, when it arrived, and what came before it. Marking a piece of writing means knowing how it was made, and that is a matter of record rather than of judgement about the finished text.

    Something the student can see too

    Feedback only teaches if the student can look at the same thing you looked at. A transcript of their own session is something to talk about together; it belongs to them as much as to you.

    It while the lesson is still running

    Anything that arrives after submission is a grade. Seeing a student stuck, or leaning on the AI, while there is still time to say something is the part that changes what they learn.

    How is AI visibility different from AI detection?

    They are not two versions of the same tool. They ask different questions, and only one of the answers has a next step attached.

     AI detectionAI visibility
    The questionMight this text have been written by AI?How did this student work?
    What you get backA probability about the finished textThe sequence — prompts, replies, revisions, in order
    Where it looksAt the artifact, after the factAt the working, as it happens
    What you can do nextRaise it with the student, or let it goTeach, mark against a rubric, or ask a better question
    What the student seesA verdict about themTheir own working, which they can learn from
    What a parent seesA number they are asked to trustThe exchange itself — what was asked, what came back
    Where it strugglesHardest on students writing in a second languageOnly covers work done inside the workspace

    How accurate are AI detectors, and who do they get wrong?

    Accuracy figures are mostly vendor-reported and conditional on how much of a document gets flagged, so the average rate is the less useful number. What matters is who the errors fall on. When researchers tested seven commercial detectors on essays by non-native English writers, every one of them misclassified that writing as AI-generated — while classifying native-speaker writing correctly.

    The likely cause is mechanical rather than malicious: a writer with a smaller working vocabulary produces more predictable text, and predictability is what these tools measure. The researchers cautioned against using them in educational settings for exactly this reason.

    In a classroom where most students are writing in their second language, that failure mode lands on the same children repeatedly. Visibility has no equivalent, because it reads what happened rather than estimating it from the prose.

    Liang et al., Patterns (Cell Press), 2023 — US eighth-grade essays vs TOEFL essays.

    What AI visibility does not do

    Worth saying plainly, because the last tool promised more than it delivered.

    It doesn't make cheating impossible.

    A student can draft somewhere else and paste it in. What the record shows is a page that arrived with no working behind it — which is a signal, but it is not a locked door.

    It doesn't decide anything for you.

    Cognity does not label a student dishonest and does not produce a score you are meant to act on. It shows what happened. The judgement stays with the teacher, where it belongs.

    It only sees its own workspace.

    Work done in another tab is not in the record. What Cognity can tell you is what happened in the task you set, which is the part you are marking.

    What teachers ask about AI detection

    Answered straight, including the ones where the honest answer is not the one we would prefer.

    See what visibility looks like, with one class.

    One task, one link, no accounts. You will know within a lesson whether reading the process beats reading a score.