# AI is eating the world’s books

An episode of [7am](https://invisiblenetwork.ai/podcast/1461999702), 27 Sep 2026, 17 min.

Audio: https://p.podderapp.com/2544644999/traffic.omny.fm/d/clips/2fb3740d-3436-44af-8cc0-a91900716aa5/62723a1f-3e96-4fc4-882c-b25e0181da4d/62e1936a-35f4-4bc5-80d5-b4cf0049fcc9/audio.mp3?utm_source=Podcast&in_playlist=70f29d90-2a63-48aa-bf8f-b25e0181da70

A rare old book can now be worth more dead than alive. Books are being bought, cut apart and scanned to train artificial intelligence. Nobody knows how many books have been destroyed this way, but second-hand booksellers around the world have reported anonymous orders ranging from thousands to as many as a million books at a time. And internal documents from AI giant Anthropic revealed plans to “destructively scan all the books in the world”. Today, author Sarah Wilson on the destruction of knowledge to build artificial intelligence. If you enjoy 7am, the best way you can support us is by making a contribution at 7ampodcast.com.au/support. Socials: Stay in touch with us on Instagram Guest: Author Sarah Wilson Photo: AAP Image/Lukas Coch See omnystudio.com/listener for privacy information.

## Transcript

[0:06] Speaker 1: That's the sound of a book being scanned to train AI.
[0:09] Speaker 1: This one remains intact. Others face a more brutal fate.
[0:13] Speaker 2: Cut the spine off it, rip the pages out, scan them,
[0:18] Speaker 2: throw it away.
[0:19] Speaker 1: Nobody knows how many rare old books have now been
[0:21] Speaker 1: bought and destroyed to feed the AI beast. But second-hand
[0:25] Speaker 1: booksellers around the world have reported anonymous orders ranging from
[0:28] Speaker 1: a thousand to a million books at a time. And
[0:32] Speaker 1: internal documents from AI giant Anthropic have revealed plans to, quote,
[0:36] Speaker 1: destructively scan all the books in the world. I'm Ruby Jones,
[0:44] Speaker 1: and you're listening to 7am. Today, author Sarah Wilson on
[0:49] Speaker 1: the destruction of knowledge to make artificial intelligence. It's Monday,
[0:56] Speaker 1: September 28th. Sarah, I thought we could begin with this
[1:04] Speaker 1: bookseller who became suspicious about who was actually putting in
[1:09] Speaker 1: orders for the rare books they owned and ultimately managed
[1:13] Speaker 1: to find out quite a lot about what was going on.
[1:16] Speaker 3: Tell me about what happened.
[1:18] Speaker 4: Yeah, so there's this kind of network of rare booksellers
[1:21] Speaker 4: and they'd been getting these large orders for rare and
[1:25] Speaker 4: unusual books, but specifically ones that were older or were
[1:28] Speaker 4: published pre-2022.
[1:33] Speaker 3: One particular bookseller.
[1:34] Speaker 4: Teamed up with a media organisation, an independent media organisation
[1:38] Speaker 4: called 404 Media. I had set up the AirTag, so
[1:42] Speaker 4: it was linked to my phone and I could see
[1:45] Speaker 4: where it was going. And they put an AirTag in
[1:49] Speaker 4: one of the books and tracked this book all across
[1:51] Speaker 4: the US.
[1:52] Speaker 3: But it left California from an airport. Next up, it
[1:58] Speaker 3: appeared in an airport in Wisconsin.
[2:02] Speaker 4: It was in.
[2:03] Speaker 3: A warehouse near the airport.
[2:04] Speaker 4: Until it finally arrived at an Amazon warehouse in Las
[2:09] Speaker 4: Vegas or the Nevada desert. And this particular warehouse has
[2:14] Speaker 4: a logo. It's a picture of a Tyrannosaurus rex attacking
[2:17] Speaker 4: a book. And what they were able to find is
[2:20] Speaker 4: that Amazon workers had been sharing information about what was
[2:24] Speaker 4: going on in this particular warehouse. They'd been charged with
[2:28] Speaker 4: removing the spine of each one of these books... so
[2:31] Speaker 4: that the pages could be easily photocopied, scanned and then
[2:35] Speaker 4: digitally uploaded. And the allegation is that these pages were
[2:39] Speaker 4: being used to feed the LLMs, the large language models
[2:44] Speaker 4: for Amazon.
[2:50] Speaker 1: And we've since learned that this isn't isolated to one
[2:53] Speaker 1: bookseller or a few booksellers or even one tech company.
[2:56] Speaker 1: It turns out that Anthropic has had this secret plan
[3:00] Speaker 1: as well to buy and then destroy large amounts of books.
[3:04] Speaker 1: So tell me about that project that they were calling
[3:07] Speaker 1: Project Panama.
[3:09] Speaker 4: Yeah, once again, it's secretive. It was only discovered through
[3:12] Speaker 4: court filings for another court case, a related court case
[3:17] Speaker 4: that Anthropic was going through for piracy. And so there
[3:21] Speaker 4: were a bunch of authors who did a class action
[3:24] Speaker 4: seeking compensation for piracy. And in the process of that
[3:28] Speaker 4: court case, it was revealed that they were also engaged
[3:31] Speaker 4: in this, what's called destructive scanning.
[3:33] Speaker 2: It's the largest copyright settlement in American history. The AI
[3:38] Speaker 2: company Anthropic forced to pay out a whopping $ 1. 5 billion
[3:42] Speaker 2: after a judge said it illegally pirated more than 7
[3:45] Speaker 2: million digitized books to train its AI model.
[3:48] Speaker 4: And the court documents that were revealed actually found that
[3:52] Speaker 4: the aim was to destructively scan all books in the
[3:56] Speaker 4: world for this particular project. But once again, Ruby, we
[3:59] Speaker 4: only found out about it by accident via another court
[4:04] Speaker 4: case with a bunch of authors, which, by the way,
[4:06] Speaker 4: I was implicated in. I am one of the authors
[4:09] Speaker 4: in this class action against anthropics. So this now seems
[4:13] Speaker 4: to be a widespread way to be able to scan
[4:17] Speaker 4: books without actually having to pay authors copyright. It's far
[4:21] Speaker 4: cheaper to buy a copy of a book, get it
[4:23] Speaker 4: shipped to some big warehouse somewhere and scan and destroy them.
[4:28] Speaker 3: And so why destroy them though?
[4:30] Speaker 1: So the books are being used to train AI models,
[4:33] Speaker 1: but why then go ahead and get rid of them afterwards?
[4:37] Speaker 4: Well, it's simply easier for them. That's the real answer.
[4:42] Speaker 4: So there is some speculation that in this court case
[4:45] Speaker 4: with Anthropic, that the fact that they destroyed them and
[4:48] Speaker 4: then replaced them with a digital copy, lended things to
[4:51] Speaker 4: a fair use defense. But I should just point out
[4:54] Speaker 4: that the court in the US did declare it fair
[4:58] Speaker 4: use to buy a book, scan it for training purposes,
[5:02] Speaker 4: and then to destroy it. The destruction is because they've
[5:05] Speaker 4: already destroyed it. They've taken the spines off. And what
[5:07] Speaker 4: can you do with a whole heap of leaves of
[5:09] Speaker 4: a book?
[5:10] Speaker 1: And you mentioned that all of these books that were
[5:12] Speaker 1: being bought and destroyed in this way, they were written
[5:16] Speaker 1: before 2022.
[5:18] Speaker 3: So tell me about why that is.
[5:20] Speaker 4: Yeah, it's significant because that was when books and materials
[5:24] Speaker 4: started to have AI content in them. And so for
[5:28] Speaker 4: AI to keep on learning, it needs to avoid recursive learning.
[5:32] Speaker 4: This is the idea where it's actually sort of like
[5:34] Speaker 4: the idea of a photocopy of a photocopy of a photocopy,
[5:38] Speaker 4: you know, 10 photocopies in. the picture is deranged and unusable.
[5:42] Speaker 4: That's what happens if AI feeds on its own content.
[5:46] Speaker 4: So to guarantee that it's not getting its own recursive content,
[5:50] Speaker 4: it has to get material from before AI was actually
[5:53] Speaker 4: introduced into the world, which is 2022.
[5:55] Speaker 2: Okay.
[5:55] Speaker 1: And so what happens then when these companies run out
[6:00] Speaker 1: of books.
[6:00] Speaker 3: To use in this way?
[6:02] Speaker 4: Yeah. Well, this is the interesting question, right? So They've
[6:05] Speaker 4: essentially scraped the internet of most of the content that
[6:08] Speaker 4: they're able to use. They're now working their way through books.
[6:12] Speaker 4: They have actually committed at this stage, I should say,
[6:15] Speaker 4: to not using books more than a hundred years old.
[6:17] Speaker 4: They're actually not that much interested in it. They actually
[6:19] Speaker 4: want modern books. And it's not even something to be
[6:22] Speaker 4: flattered about that my books were used because they're actually
[6:25] Speaker 4: really kind of trying to get hold of weird like pamphlets,
[6:28] Speaker 4: like the sewerage techniques of greater Chicago region in the 1980s.
[6:31] Speaker 4: Like They're wanting really obscure stuff because they're wanting to
[6:35] Speaker 4: extend the learning capacity of their technology. But when they
[6:40] Speaker 4: run out, that is a very good question. And it
[6:42] Speaker 4: looks like they're going to run out in about 2028.
[6:44] Speaker 4: That's what the modelling is suggesting. After that, they can
[6:47] Speaker 4: actually have to turn to real life authors and writers
[6:51] Speaker 4: to create more content, presumably. So it's a very real
[6:55] Speaker 4: risk because when they do run out, their AI actually
[6:58] Speaker 4: kind of starts to grind to a halt. And I
[7:00] Speaker 4: think it's quite ironic that they're going to need writers. Now,
[7:03] Speaker 4: this has also happened with philosophers. Philosophers are now earning
[7:07] Speaker 4: sort of six-figure incomes working for these tech companies because
[7:12] Speaker 4: they're needed to navigate the moral maze that the technology
[7:16] Speaker 4: has created for humankind. And it's quite interesting, isn't it,
[7:21] Speaker 4: that it's actually these thinkers, writers, artists, creatives... The people
[7:27] Speaker 4: are the very thing that AI can't create and is
[7:29] Speaker 4: going to have to rely on going forward.
[7:36] Speaker 1: Still to come, why Sarah accepted money from Microsoft to
[7:39] Speaker 1: use her books for AI training.
[7:49] Speaker 3: So you mentioned... that you were part of this case, Sarah.
[7:52] Speaker 1: So before Anthropic was buying up these books, it was
[7:55] Speaker 1: using these pirated.
[7:56] Speaker 3: Online libraries to train its models.
[7:59] Speaker 1: Your books were part of those libraries, and then you
[8:01] Speaker 1: were part of this group of authors who ended up
[8:03] Speaker 1: settling with Anthropic for, it was $ 1.
[8:05] Speaker 3: 5 billion.
[8:07] Speaker 1: I mean, when you first found out that Anthropic was
[8:10] Speaker 1: using your books in that way, how was that realization
[8:13] Speaker 1: for you?
[8:14] Speaker 4: I wasn't shocked, to be honest, because I'd been witnessing
[8:16] Speaker 4: this kind of thing happening over the years. My content
[8:20] Speaker 4: had been used illegally without my authorization for all kinds
[8:24] Speaker 4: of things. There's publishers in Iran, for instance, that take
[8:28] Speaker 4: my books, scan them, translate them and produce them as
[8:31] Speaker 4: physical books and just put them out to market and
[8:33] Speaker 4: make money from them. There's nothing I can do because
[8:35] Speaker 4: they're not part of international copyright laws. I know that
[8:40] Speaker 4: my work has been all across various LLMs. for many
[8:44] Speaker 4: years because people can go in there and say, please
[8:47] Speaker 4: write this in the style of Sarah Wilson, and it
[8:49] Speaker 4: spits out literally the kind of phraseology that I've been
[8:52] Speaker 4: using for years.
[8:53] Speaker 3: So it's in there.
[8:55] Speaker 4: Authors can also go into some of these pirate libraries
[8:58] Speaker 4: and do a search and can find their work in there.
[9:00] Speaker 3: So it wasn't a surprise.
[9:02] Speaker 4: There's also about 75 cases that are actually out there
[9:06] Speaker 4: at the moment, various authors suing other companies. So this
[9:11] Speaker 4: is just the first one to come through with a settlement.
[9:14] Speaker 4: And it's also the largest in history. Okay.
[9:18] Speaker 1: And after that class action, you were invited to opt
[9:21] Speaker 1: into a licensing deal with Microsoft, which would allow them
[9:25] Speaker 1: to use your books.
[9:26] Speaker 3: To train AI.
[9:27] Speaker 1: There were some authors who refused to do that on principle.
[9:32] Speaker 3: You decided to do it, to take the fee, though.
[9:34] Speaker 3: Tell me about that decision.
[9:35] Speaker 4: Yeah, I did think about it for quite some time
[9:38] Speaker 4: and the bulk of authors did take the fee. And
[9:40] Speaker 4: I share their thoughts, but this issue is massive. There
[9:45] Speaker 4: are multiple LLMs out there who are taking our content anyway.
[9:50] Speaker 4: So we should take whatever money is being put forward.
[9:53] Speaker 4: People should be paid. We get paid licensing fees if
[9:56] Speaker 4: somebody wants to use our work for a PowerPoint presentation
[10:01] Speaker 4: or in schools for lectures and this kind of thing. Now,
[10:04] Speaker 4: of course, the devastating issue is that the books aren't
[10:07] Speaker 4: just being used by university students in a lecture. They're
[10:11] Speaker 4: being used to train students a system that could replace writers, artists, creators.
[10:18] Speaker 4: That is the bigger issue here. Now, is it going
[10:20] Speaker 4: to be fixed by authors refusing to take the money
[10:23] Speaker 4: that they can? You know, there are principled decisions that
[10:27] Speaker 4: could be made here.
[10:28] Speaker 3: Ultimately, I decided to.
[10:29] Speaker 4: Take the money because I sort of saw it and went, okay,
[10:32] Speaker 4: I'm taking money while I can, and I'm going to
[10:35] Speaker 4: put it into funding my ability to go and fight
[10:37] Speaker 4: the bastards.
[10:38] Speaker 3: And so how do you do that then? Once AI
[10:40] Speaker 3: has gone.
[10:41] Speaker 1: Through all of these books, used all this knowledge, its
[10:44] Speaker 1: language skills have improved as a result, how do you
[10:48] Speaker 1: fight back to make sure that writers aren't made obsolete
[10:52] Speaker 1: by AI?
[10:53] Speaker 4: Yeah, there's a couple of things. I mean, I'm talking
[10:55] Speaker 4: to you right now and exposing this. We're having a
[10:58] Speaker 4: conversation about it so more and more people know about it. Remember,
[11:01] Speaker 4: we became aware of this because some rare booksellers started
[11:05] Speaker 4: talking to each other and decided to expose it, right?
[11:08] Speaker 4: This topic is now being talked about around the world.
[11:11] Speaker 4: the European Commission has had discussions about copyright laws. Here
[11:15] Speaker 4: in Australia, the Australian Society of Authors have also got
[11:18] Speaker 4: together and are lobbying the government and have done so successfully.
[11:22] Speaker 4: They've managed to ensure that there are copyright protections in place,
[11:26] Speaker 4: that they're not destroyed. And they're also pushing for ways
[11:29] Speaker 4: to get compensation for authors going forward, but also for
[11:31] Speaker 4: past compensation. So I think that is part of what
[11:36] Speaker 4: needs to be done right now. We should keep in
[11:38] Speaker 4: mind that, you know, These tech leaders love to talk
[11:42] Speaker 4: in very, very dramatic language. We've seen a lot of
[11:45] Speaker 4: that happen in the last week or so with the
[11:47] Speaker 4: whistleblowers coming out. You know, so when they say that
[11:50] Speaker 4: they're destructively scanning all books in the world, they're in
[11:54] Speaker 4: fact destructively scanning one copy of each book. It's not
[11:57] Speaker 4: like they're taking all copies of my books and putting
[12:00] Speaker 4: through an incinerator, but they are actually taking the content,
[12:04] Speaker 4: the creative content. The other thing I would say is
[12:07] Speaker 4: this is actually a sector of the broader debate that
[12:11] Speaker 4: can be legislated.
[12:13] Speaker 2: We can actually control this.
[12:14] Speaker 4: We've seen it with the music industry, the Napster case
[12:18] Speaker 4: over 20 years ago. That actually revolutionised the music industry
[12:22] Speaker 4: and provided copyright provisions for artists. We can do the
[12:26] Speaker 4: same with books. It's not difficult to do. It can
[12:30] Speaker 4: be legislated against. So the more that people know about this,
[12:33] Speaker 4: the more pressure is put onto governments around the world
[12:36] Speaker 4: I do believe something will come out of this exposure.
[12:40] Speaker 1: And as you've noted, AI companies are now hiring philosophers
[12:44] Speaker 1: to help them navigate this problem that they've created.
[12:48] Speaker 3: You've raised the prospect.
[12:49] Speaker 1: Of writers being hired when all of the books run out.
[12:52] Speaker 1: It seems back to front though, doesn't it? To create
[12:55] Speaker 1: the problem by pursuing AI and then try and get
[12:57] Speaker 1: human thinkers back in to solve it. What kind of
[13:01] Speaker 1: future do you think we are hurtling towards here?
[13:04] Speaker 4: We are hurtling to a very uncertain future. And I
[13:08] Speaker 4: think what's been exposed in the last couple of weeks,
[13:10] Speaker 4: what a lot of people are becoming aware of is
[13:12] Speaker 4: how little the tech bros themselves know about what that
[13:17] Speaker 4: future looks like. You know, they're very much stabbing in
[13:19] Speaker 4: the dark. They're asking to be regulated for pauses to
[13:23] Speaker 4: be put on the development of their technology. And the
[13:26] Speaker 4: best thing that we can all do is be very,
[13:27] Speaker 4: very aware of what's happening and And to use whatever
[13:30] Speaker 4: democratic processes still remain in place to fight for, I guess,
[13:36] Speaker 4: retaining things that are so, so, so important to sort
[13:40] Speaker 4: of human life. And I guess the worst outcome for
[13:44] Speaker 4: us going forward is something that's totally Orwellian. George Orwell's 1984,
[13:50] Speaker 4: The Mystery of Truth, wound up controlling the one document
[13:55] Speaker 4: of recorded history and was able to change it at
[13:57] Speaker 4: will to suit its party lines. Now, I don't think
[14:00] Speaker 4: we're close to that yet. There's a lot of books
[14:03] Speaker 4: out there in the world. And so far, as I say,
[14:06] Speaker 4: the tech companies are actually committing not to buying up
[14:09] Speaker 4: books that are over 100 years old. We've got to
[14:11] Speaker 4: ensure that stays in place. So yes, there's a dystopian
[14:15] Speaker 4: future ahead of us if we don't act now I
[14:18] Speaker 4: mean how many times have we heard that but I
[14:19] Speaker 4: think actually this is actually viscerally disturbing anybody who's seen
[14:25] Speaker 4: the imagery of these books with their spines being cut
[14:28] Speaker 4: off it's actually really upsetting because books are so precious
[14:32] Speaker 4: to us people preserve books we create massive monuments to
[14:36] Speaker 4: books for preserving our history, 5,000 years of mindful recording
[14:42] Speaker 4: of the human moral endeavour. And so I do think
[14:45] Speaker 4: that this particular corner of the AI travesty, we probably
[14:50] Speaker 4: will be able to ensure that it doesn't go too far.
[14:53] Speaker 1: Yeah, books are text, but they're also symbols in a
[14:56] Speaker 1: lot of different ways.
[14:57] Speaker 3: Absolutely. Sarah, thank you so much for speaking with me.
[15:02] Speaker 3: My pleasure, Ruby.
[15:02] Speaker 1: Thank you. Also in the news, the federal government says
[15:22] Speaker 1: it's working closely with OpenAI after one of its AI
[15:25] Speaker 1: agents gained unauthorised access to Australian government websites. OpenAI boss
[15:30] Speaker 1: Sam Altman and Anthropic CEO Dario Amodei have now been
[15:34] Speaker 1: called to appear before a Senate inquiry following similar incidents
[15:38] Speaker 1: involving government sites in Australia and the United States. Prime
[15:41] Speaker 1: Minister Anthony Albanese says OpenAI has questions to answer after
[15:46] Speaker 1: it took almost three months to notify the government about
[15:49] Speaker 1: the Medicare breach. And Iran says diplomacy is the only
[15:53] Speaker 1: way forward with the United States after President Donald Trump
[15:57] Speaker 1: rejected its latest proposal to reopen the Strait of Hormuz.
[16:01] Speaker 1: Tehran had offered to reopen the shipping route within seven
[16:03] Speaker 1: days if the US eased military pressure and lifted its
[16:06] Speaker 1: blockade on Iranian ports. Iran says it's still waiting for
[16:10] Speaker 1: a formal response through mediators, while Trump has called for
[16:13] Speaker 1: Iran's continued economic isolation.
[16:16] Speaker 3: I'm Ruby Jones. This is 7am. Thanks for listening.

Shows are listed from Apple Podcasts' charts and directory, and Podcast Index where it is used. Every show and episode belongs to its publisher and plays from the publisher's own feed; this site hosts none of them.