Context · AI
Schools are running work through tools that claim to spot AI writing, and those tools are wrong often enough in both directions that a result should never be the end of a conversation at home.
They flag real writing as fake
These detectors look for text that is too even and too predictable, which is also what careful writing by a second-language student or a kid who was taught to write in clear short sentences looks like. Honest work gets flagged regularly, and the student has almost no way to prove a negative.
They miss AI that has been nudged
Asking ChatGPT to write in a more casual voice, or editing a few sentences by hand, is usually enough to get past a detector. So a confident pass tells you very little either, and a family that treats the app as proof in either direction is trusting something that does not work.
What actually tells you
The reliable test is not another app, it is a conversation. A kid who wrote it can explain why they made a particular choice and what they nearly wrote instead. A kid who had the AI generate it usually cannot, and that takes about ninety seconds to find out.