I'm not sure what you mean by this. Your current solution passes, and yes, the two strings have to have the same quantity of each letter, otherwise they aren't anagrams.

edit: nvm, I see. You're saying that your kata passes the tests, but not this specific case. Yes, this case should be added as a fixed test and the random tests should also be guaranteed to generate at least one random test case like this.

I just tried and AI returned me working solution immediately, and I am not really sure what is the problem here, I'm sorry :(

But I admit that things like recursion and type checking might put it in the range of harder 7 kyus. Maaaaaybe the low end of 6 kyu would not be wrong either. But even then, I don't think that 7 kyu rank is very wrong, and rather justified.

What is the point of using AI in a service whose sole purpose is to train programming?

And you can't even brag because your own comment gives it away...

Oh come on, I said literally zero words about me to show off how clever I am :(

I'm rather surprised to learn that AI can't handle anything above an 8 kyu problem.

Well, there is 3207 solutions from humans which did work. Maybe try writing one by yourself?

I would also like to remind everyone interested that submitting AI generated code as solutions is agains Code of Conduct.

I am very curious how your solution passes tests on your PC with the same data if it does not even solve the right problem?

Your solution would work with regular Python lists (well, almost, it would fail for some cases), but it will not work for generic linked lists.

Your solution tackling wrong problem is not a valid kata suggestion.