Some safeguards have been added which curtail certain direct misbehavior, but it is still capable - by your own admission - of doing it. And it still profits from the unlicensed use of copyrighted works by using such material for its training data. Because what it is producing is not a new and unique creative work, it is a composite of copyrighted work. That is not the same thing.
And if you are comparing LLMs and hammers, you’re just proving how you fundamentally misunderstand what LLMs are and how they work. It’s a false equivalence.