I still cannot imagine how IA thought that giving unrestricted access to copyrighted books was a good idea. It seemed inevitable that someone would sue them over it. Honestly, I think that IA's ambivalence towards the use of their website for outright piracy might lead to their collapse, and that's a shame. The Archive can be a really wonderful tool, though I'm not sure that its current management really knows what t…
IA was trying to act like a library: they bought physical books and lent digital scans of them, ensuring that only one user can read one book at a time. So IA's position is that you can treat digital books like physical books, i.e. re-sell them or lend. The only difference is that they don't require you to come to library in person.
Publishers position is that digital books are different from physical; you have no right to re-sell or lend it without publisher's permission. This is what this case is about.
IA's arguments that this was fair use seem exceptionally weak: * Their "transformation" argument comes down to easing access, which has already been shot down in previous cases; unsurprising, since Napster could make the same argument. * Their "nature of the work" argument came down to the fact that some of the books they scanned were nonfiction. * They made a halfhearted attempt to claim that "amount and substantial…
If you buy a physical book, you have a right to lend it for free or re-sell it (first-sale doctrine). The publishers want to establish a precedent that you cannot do the same thing with digital books: the library must buy a special license and pay for every reading.
So this is about stripping people from their rights regarding to books made with a new technology.
> IA will let you download full video games for the switch that are still being actively sold [1] I am not seeing that anywhere. I see a file called “My Nintendo Switch games collection” and it is a big jpeg photo of a bookshelf. Is this what you mean?
It's harder to find them than I expected, but one search with a result is "super mario wonder nsp". I had the advantage of knowing the format(s) Switch game dumps come in, though, so the average person might not find much.
I’m getting a “no results found” response from that search. Maybe it has already been taken down?
I hope this doesn't bankrupt the Internet Archive (either the legal fees or the case - I don't recall what they're asking for). It would be bad if the Wayback Machine, the biggest internet archivist around, went under, and also all the books, software, et cetera that the Archive hosts. I wonder if there's any way to archive all of the Wayback Machine (82.3 petabytes), or, better yet, all of the Internet Archive (whic…
For wayback machine, are those compressed, deduplicated numbers?
A semi-popular domain can have millions of results on their CDX api, but with https/https duplicated and about 90% of results are error pages or pages with deliberate garbage / LFI attempts in them.
To me, the point isn’t that what the IA was doing was fair use, but that what LLMs are doing arguably is not . > In the AI use case, they're typically aiming not to output any significant part of the training data What they’ve aimed to do and what they’ve done are two different things. Models absolutely have produced output that closely mirrors data they were trained on. > not competing in the market with the origina…
But look at your own argument. LLMs are not fair use because they might be prompted into regurgitating something substantially similar to the trained data. And yet, the IA is 100% aiming to absolutely reproduce literally every part of the work in a 100% complete manner that replaces the original use of the work. And you cannot bring yourself to admit that the IA is wrong. When you get to that point you have to admit…
IA was lending a digital copies (only one user at a time may read the book), it was acting like a library lending out physical books, only IA did it over the Internet which is more convenient. IA is non-profit.
What publishers argue is that you cannot treat digital books like physical ones; i.e. you cannot re-sell or lend (like IA did) a digital book.
What LLM do is that they use copyrighted content for profit and do not lend anything.
This has been playing out for many years. And it's all because Brewster Kahle decided that an overly broad interpretation of the Internet Archive's mission trumped the rights of authors and publishers, and the laws of the United States. When IA was asked to stop CDL - many times - he continued. The National Writers Union tried to open a dialogue as early as 2010 but was ignored: The Internet Archive says it would rat…
If all information wants to and should be free than how can freeing information ever be wrong?
> If that were the case then each copy would need to sell for thousands of dollars for content creators to afford food. We have an enormous surplus of content creators and most of the content is not very good. I don't see why we need to structure our economic system such that people must be able to making a living churning out mediocre scifi/romance/mystery novels. If they can, great, but I don't think that's the goa…
> I doubt it's really much more than the number of sales now Yikes. I can not possibly disrespectfully disagree more with everything you said. Baldur's Gate 3 has sold about 15 million copies. It's peak concurrent user count on Steam is 875,343. A difference of about 20x that will continue to grow as BG3 will sell meaningful copies over the next 10 years. Limiting sales to peak CCU is categorically insane. And deeply…
The people that would borrow the game from the library to play it might at best pirate it if they couldn’t get it from the library. Maybe they’d pay a few bucks tops rather than $60-80. Library game borrowers are not big game buyers in the first place.
Games can and do already get around this anyway via software, if you want to argue the laws should work differently for them then I’m open to that, but I also don’t think games matter enough on a societal scale that we should tolerate current copyright laws in order to protect video game studios over the long tail of disappearing orphaned works.
Something like this would never have fallen under fair use in the first place, I don't think. It doesn't really mesh very well with any of the factors US law lists as considerations when deciding whether or not use is fair. If this falls under anything it falls under the first sale doctrine. I generally side with publishers and artists on the generative AI debate, but I'll at least concede that they have some grounds…
A physical library does not involve making a copy, but lending out an existing physical copy. One book, one reader. "Controlled Digital Lending" literally involves making and transmitting a copy to another physical medium electronically. Brewster has endangered the good work done by people at the archive for a case he was never going to win and which was of dubious value. If he wants to be the next Anna's Archive the…
1) Making a copy is not always illegal; for example, scanning books for Google Books by Google is considered legal. 2) Digital books are DRM-protected so you cannot lend them or re-sell like you can with physical books. So making a digital copy of a physical book can be considered merely a method to workaround these restrictions. 3) Publishers want to use new technology (electronic books) to remove rights that consum…
> 1) Making a copy is not always illegal; for example, scanning books for Google Books by Google is considered legal.
I believe I remember this was tried at one point too. The significant difference here is that Google has seemingly unlimited money to throw towards lobbying and court cases
Copying text out of the PDF from my phone is not fun! The gist is that since the CDL isn't okay then they don't really need to deal with NEL as it's predicated on the legality of the CDL (pages 15-16). The NEL gets a couple of sentences, the bulk of the ruling is about the CDL
Wasn't NEL the basis of the original suit?
Yes! The U.S. Court of Appeals upheld the prior 2023 ruling against IA, the case stemming from the IA's National Emergency Library (NEL) initiative during the COVID-19 pandemic, which allowed users to access digitized books without the usual lending caps.
This prompted the lawsuit from several major publishers, who argued that this violated copyright law.
The appeals court rejected the IA's argument that its activities fell under the "fair use" doctrine, specifically noting that the IA digital library acted as a substitute for original books, depriving publishers and authors of revenue.
While the court acknowledged that the Internet Archive's activities were non-commercial in nature, it still concluded that the wide availability of digitized books due to unregulated replication harmed the market for the original works.
This decision has significant implications for the IA’s future operations, potentially limiting its ability to continue its broad digitization efforts without publisher consent.
The ruling reinforces the legal rights of publishers and authors to control the reproduction and distribution of their works, even in digital formats.
The Internet Archive has few remaining legal options, with the Supreme Court being one of its last possible avenues for appeal.
Meanwhile, the organization faces additional lawsuits related to its music digitization efforts, those litigations are ongoing.
The ruling today highlights a broader conflict between the rights of creators and the push for wider public access to information, with the court siding firmly with the former. The case sets a strong precedent for how copyright law is applied to digital libraries in the U.S. moving forward.