I teach to first year students. We have graded programming assignments they can do by themselves, where they can cheat in all sorts of ways, but I honestly don'

I teach to first year students. We have graded programming assignments they can do by themselves, where they can cheat in all sorts of ways, but I honestly don't care, it's their responsibility to learn. Then we have theoretical exams, where there are also questions about programming (but we don't check if the code has perfect syntax). In one course the exams are oral: excellent tool for assessing students, but very time consuming, in another it's written: more efficient, but more shallow. I like the idea about teaching how to use AI, but it has to be a tool among others.