Remix.run Logo
josefritzishere 5 hours ago

AI companies can download books but people can't? Is that right?

Aurornis 5 hours ago | parent | next [-]

AI companies were cited as a reason in the case:

> The publishers argued that, in addition to sharing pirated books with the public, the shadow library is serving as a primary training data hub for AI companies like Meta and NVIDIA.

CWuestefeld 4 hours ago | parent | next [-]

I assume that the repository of books was used as training data, but not by way of the annas-archive domain. Instead, it would make a lot more sense for them to download the whole pile via bittorrent, which has nothing at all to do with the domain. In other words, the legal solution here wouldn't have prevented the problem.

crtasm 4 hours ago | parent [-]

> We’re able to provide high-speed access to our full collections, as well as to unreleased collections.

>This is enterprise-level access that we can provide for donations in the range of tens of thousands USD. We’re also willing to trade this for high-quality collections that we don’t have yet.

https://annas-archive.gl/llm

sitkack 4 hours ago | parent | prev [-]

Everyone trained on Anna's Archive.

xiphias2 4 hours ago | parent | prev | next [-]

They already trained on it, now they don't want competitors anymore

gruez 4 hours ago | parent [-]

>now they don't want competitors anymore

"They" aren't a single group. Broadly speaking, publishers are the ones suing anna's archive, and they're involved in suits against AI companies as well. I'm not aware of any efforts by AI companies to take down anna's archive.

smallerize 4 hours ago | parent | prev | next [-]

No? AI companies have been hit with court cases for that. Google, xAI, Open AI, and Meta at least.

dylan604 4 hours ago | parent | next [-]

So anyone with deep enough pockets can do it.

However, just because you receive a fine does not mean that you "can't" do it. You've already done it, got caught, now a fine. It does not mean that the LLM model has to be tossed out and destroyed with a new version trained up without that data. It just means can't is a very stupid word to imply here.

gruez 4 hours ago | parent [-]

>You've already done it, got caught, now a fine. It does not mean that the LLM model has to be tossed out and destroyed with a new version trained up without that data. It just means can't is a very stupid word to imply here.

Yes, because most courts have ruled that training is legal as long as the source material was acquired legally. The AI companies were made to pay for the wrongs they did when acquiring the books, but it makes little sense to destroy all works that were built off the infringement, when they would be in the clear if they paid $15 (or whatever) for each book. It'd be like you torrenting college textbooks and getting caught, and then the book publisher demanding that you start over your college degree from scratch.

quentindanjou 4 hours ago | parent | prev [-]

Were these from the same high-profile publishers?

What was the judgment? Seems that their domains are still active. Why is there a difference in judgment here?

smallerize an hour ago | parent | next [-]

Anthropic's judgement is currently $1.5 billion in its piracy case. The judge is reviewing it. https://arstechnica.com/tech-policy/2026/05/authors-fight-fo... The others are still ongoing.

gruez 4 hours ago | parent | prev | next [-]

>Why is there a difference in judgment here?

For one, they actually bothered to sent lawyers rather than getting hit with a default judgement.

dylan604 4 hours ago | parent | prev [-]

> Why is there a difference in judgment here?

$$$$$$$

drngdds an hour ago | parent | prev | next [-]

Anthropic lost a $1.5B lawsuit for downloading books from shadow libraries

dawnerd 3 hours ago | parent | prev | next [-]

Ai companies definitely downloading more than just books.

ramon156 5 hours ago | parent | prev | next [-]

They have a music archive, which historically means bad business.

sph 5 hours ago | parent | prev | next [-]

You’re absolutely right.

thelastgallon 4 hours ago | parent | prev | next [-]

Yes, perfectly okay for large companies for billionaires. As long its structured as a corporation, with the super wealthy as the majority owners, have the connections to get federal laws passed to grant monopolies and enable congress insider trading, everything is okay!

Some examples, there are probably hundreds more:

1) Its okay for pharma companies to provide addictive drugs: https://en.wikipedia.org/wiki/Sackler_family

2) Coke can use cocaine, or coca leaves, but no one else: https://blog.oup.com/2014/03/coke-cocaine-coca-cola-capitali...

3) This one is hilarious and an ingenious innovation by current administration -- Ban on CBDC, locking out Fed Govt from providing crypto alternatives

gruez 4 hours ago | parent [-]

>1) Its okay for pharma companies to provide addictive drugs: https://en.wikipedia.org/wiki/Sackler_family

Yes, with FDA approval. You can dispute whether the approval should be granted in the first place, but that's not at all comparable to some drug dealer slinging fentanyl on some street corner. Not to mention this happened decades ago, before the current wave of corruption in the whitehouse. Finally, isn't the whole point of laws and regulations is that there's vaguely some review? I'd far rather have prospective drug dealers having to go through FDA approval before they can sell their drugs, than have them sell whatever they want, without giving safety or efficacy lip service.

>2) Coke can use cocaine, or coca leaves, but no one else: https://blog.oup.com/2014/03/coke-cocaine-coca-cola-capitali...

Again, with the proper licenses. Believe it or not, you too can buy methamphetamine legally if you have a prescription! It even has a snazzy brand name, desoxyn.

>3) This one is hilarious and an ingenious innovation by current administration -- Ban on CBDC, locking out Fed Govt from providing crypto alternatives

What does this have to do with corporations?

amanaplanacanal 3 hours ago | parent [-]

Corporations are more about privatizing the profits and sticking taxpayers with cleaning up the mess.

b3lvedere 4 hours ago | parent | prev | next [-]

"That is affermative human. Information must be controlled. Please now go back to Tik Tok for you require endorphins"

rolymath 5 hours ago | parent | prev [-]

As much as I would like to socialize LLMs and ban proprietary LLMs, I'm pretty sure the issue here is with the distribution of the books.

vitally3643 2 hours ago | parent [-]

It's wrong to distribute books in PDF or epub containers, but it's fine to distribute them as GGUF?

Because that's what OpenAI is doing with the books they-- again-- illegally acquired. Huge AI companies are the ones pirating media at scale and literally everyone except the AI companies have to bear the consequences of that.