Here is a relevant quote from the TFA:
> Although students were permitted to use AI to brainstorm topics and identify sources, in this instance the students had indiscriminately copied and pasted text from the AI application, including citations to nonexistent books (i.e., AI hallucinations).
(And re better rating: sadly making classes less useful usually improves the rating. I sure if there was a class where they were just watching cartoons, with super-simple quizzes at the end that everyone could answer, it'd have highest ratings from most students, as well as high ratings from many parents, and best scores on finals. Now, it might not work that hot in real world, but by that time the school would be long over...)
Why do you expect children to know math only after months of tries, but understand the perils of AI after hearing one sentence regulation? That's not going to help the kids. You need to spend time practicing using the tool with them, showing them the pitfalls in practice, and only after enough time you can start rating them.
The school handed them a gun and they're Pikachu surprised a kid got shot, and now they're blaming it on the kid that had it in hands - but the school is to blame. And it's certainly not newsworthy, or where is the math exam results article?
What you are describing might make sense for "ChatGPT class", but that wasn't it, that was AP History.
(And that's how schools work in general: in real life, no one integrates matrixes by hand; and yet calculus classes do not teach CAS systems or their perils)
Remember the "disciplinary action" here is giving the student a bad grade, with the opportunity to redo the assignment.
Are you seriously asserting they should've gotten a good grade for a paper that cites sources that don't exist? In an advanced placement class no less?
If anything they're getting of lighter than students who did a bad job on their own without using AI. I know I was never given a chance to redo a paper I phoned in.