The Pros and Cons of TDM Text and Data Mining

There are two parallel debates happening right now. One is the legality of commercial and research entities scraping large amounts of data from the web in order to train machine learning models. The other is whether exactly AI produced works can count as, you know, like authored works in a traditional copyright sense. What have you found and kind of what you're feeling on this without obviously being able to be definitive because it's quite complicated?