So I've used Suno a bit in my own musical creations, but would never release something with any sound that was generated via AI. What I do is write and record a song (I'm also in Ableton), then when I'm unhappy with something or don't know where to go next, I throw it into Suno and have it generate some ideas. I use those ideas just as you would use ideas from a writing partner. If I like them, I transcribe them, play them on real instruments, and record them. So no actual sound from Suno ever makes it into my compositions - just some of the musical ideas here and there. That way it's still me playing and still sounds like me.

To me, this is the most legitimate way for music makers to use AI. If you go look at the credits for almost any professional recording artist, they all work with one or more writers and producers to get their music dialed in. Us normies can't really afford to have Max Martin come over and write with us, and I think using AI in the way I mentioned is a suitable alternative. I really fail to see the difference, honestly.