Nothing stops you from downloading Ann’s archive and training a model on it, right? The likelihood that you, as an individual, get sued over is is virtually zero.
This is what Meta tried to do, quietly download and use the data, to do research and advance their LLMs, without trying to establish any legal precedents or pick up fights.
In Germany people are sued for illegal movie downloading all the time. It's hard to imagine the companies behind that operation aren't aware that you can also download books.
This is what Meta tried to do, quietly download and use the data, to do research and advance their LLMs, without trying to establish any legal precedents or pick up fights.