How, to you, does training an AI violate copyright?
Otherwise AI becomes a glorious copyright-laundering machine, use something that you want to copy to train an AI, then ask that AI to replicate the stuff it was trained on. Of course it will make a perfect copy, but who is the creator of that intellectual work? The AI itself not for sure, but the original author of the stuff that the AI copied.
An no, it doesn't matter if the AI doesn't copy exactly but does a bit of shuffling of stuff, if not you are arguing that it's not cheating if you copy the homework of your classmate but you rearrange some words to make it not identical...
From copyright.gov "Copyright does not protect facts, ideas, systems, or methods of operation, although it may protect the way these things are expressed"
What rights does copyright provide?
Reproduce the work in copies or phonorecords.
Prepare derivative works based upon the work.
Distribute copies or phonorecords of the work to the public by sale or other transfer of ownership or by rental, lease, or lending.
Perform the work publicly if it is a literary, musical, dramatic, or choreographic work; a pantomime; or a motion picture or other audiovisual work.
Display the work publicly if it is a literary, musical, dramatic, or choreographic work; a pantomime; or a pictorial, graphic, or sculptural work. This right also applies to the individual images of a motion picture or other audiovisual work.
Perform the work publicly by means of a digital audio transmission if the work is a sound recording.
Training an AI isn't on this list. An AI may produce infringing outputs, but training an AI isn't a violation of copyright.
"Copyright laundering" is a nonsense idea. Copyright is not designed to prevent people or machines from producing things in a style or similar things. It protects against exact reproduction or reuse, and that only to the extent that the form is separable from function. If the form is inseparable from function, the expression cannot be copyrighted at all.
Let's assume "shuffling stuff" as you describe it would solidly place a work into the "derivative works" category.
Fine, the AI has produced a derivative work. The original copyright holder has a claim on that work if it were to be distributed, sold, displayed, or performed.
Does the AI hold a copyright? No. Only humans can hold copyrights. Does the prompter hold a copyright? No. A human has to perform the transformative work, not merely arrange for a machine to do it. AI outputs are not copyrightable.
But how you "feel" about it isn't a useful measure. It being like "cheating off someone's homework" isn't useful, either. Hell. Copyright is carefully crafted to allow "cheating off someone's homework" provided that the result doesn't reproduce elements of the original exactly.
I am also of the opinion that AI weights aren't copyrightable, as they're the result of an automatic process. "Open source" weights are a laughable non-sequitur.
I am not generally of the opinion that we need to expand copyright in order to protect a status quo.
How it's not? Copyright means that you, the author, owns the rights to a software, a song, a book, etc. and you give permission in form of what is usually formalized in a license to other people to use your copyrighted work in some ways. You decide what people can do with your copyrighted material, and what it's not expressively allowed in your license it's not permitted. Moreover everything you create (even this comment!) as by default a copyright, and if you don't say anything the copyright is intended to be exclusive (i.e. you can't do anything with it).
Now, unless your license doesn't expressively say that you are granted the permission to use that copyrighted material to train an AI model you can't. To me it's clear, I don't know how someone could think of the opposite. What copyright has for example a webpage? If it doesn't have a license associated with it that is compatible with scrapping it and training an AI model that operation is illegal.
Otherwise you are saying that it's a crime for me to download a book from the internet, but it's not a crime to take such book and use it to train an AI model then then has all the knowledge that is present in that book? Does make sense to you?
Of course there is the argument of "fair usage", but could we consider "fair" taking systematically entire software to train on it AI models?
And before you say "but humans can study a software", it's because the license of such software allows it, but the key is that it allows humans, in no way you could for example state that the GPL license allows it to be used as a training source for AI models without the consequence of any of the output of such AI models being considered derivative work.
You can read about it right here: https://www.law.cornell.edu/uscode/text/17/106
Those protections and no other does copyright afford.
You are describing your feelings, not a law.
Also, it probably is a crime to download a book from the internet (or you could live in Switzerland). The owner/operator of the server has pretty certainly committed copyright infringement. Merely receiving a copy doesn't constitute infringement. In civil court they could make the case that you caused the copy to be made by clicking.
What you do with the book cannot be controlled, save reproducing it.
But there's the crux. Does placing the words into RAM qualify as "reproduction"? The ninth circuit held that it did in MAI Systems vs. Peak Computer. The second circuit held that it did not in Cartoon Network v. CSC Holdings.
The supreme court will have to decide.
Does that violate copyright? What if we us the binary source code of MacOS, Linux and Windows, then ask it to recreate Windows?
Does the fact that we added some other ingrediants violate the idea that the LLLM will create a substantially similar product?
I'm not defending windows, but indicting LLMs as anything but "fair use".
source: online learning of codebases in the hundreds of thousands to millions of lines of code