Amazon and the Dystopian Future of Book Censorship

If you believe that paper books are free from internet censorship, think again. Books aren’t hand copied by monks anymore, they’re printed from a digital source, in just a handful of places. This digital format is what allows censorship to take place just before a book hits the physical press.

In the process of trying to get my own book distributed, I experienced a glimpse of this dystopian future. In my case it pertained to a trademark or copyright issue, but the same content scanning techniques can be deployed to e.g. screen for terrorist content, bad opinions, inconvenient facts, etc..

My book is called Bitcoin: A Work in Progress: Technical innovations from the trenches. It’s based on the Bitcoin, Explained podcast that I host with Bitcoin Magazine technical journalist Aaron van Wirdum. Each chapter of the book is based on an episode and explains some interesting technical aspect of Bitcoin.

I initially planned to distribute this book through Amazon KDP, which offers 60% royalties on paperbacks, and they have useful tooling. One feature that’s particularly helpful is the ability to order proof copies. This allowed me to iterate on the cover, font size, page margins, QR codes, etc. and see what it really looks like. For €6 they print the book in Poland, throw it on truck across Germany and it gets to my home in The Netherlands in just a couple of days.

What Happened?

Once the book reached a nice thickness, I pressed the Publish button, expecting the book to be live just in time for the Bitcoin 2022 conference in Miami. Instead, I ended up in a kafkaesque dialogue with either a bot pretending to be support staff, or support staff pretending to be a bot.

This immediately raised the question: how can someone file an infringement claim against a book that doesn’t exist yet? Obviously this is an automated system that went through the cover and content of my book and found something that it believed was an infringement. There’s no actual claim.

In any case, the demand is ridiculous: I have to find an author who doesn’t exist, because I wrote the book. But I’m not allowed to simply state that I wrote the book. They might as well demand a shrubbery.

So I replied to the email asking for clarification. After a full week I received the following “reply”:

Again, notice the absurdity. There is no “edition previously published on Amazon”, which is obvious when you search for the title on Amazon. I included a screenshot of that search in my reply.

Meanwhile I found an article What to Do if Amazon KDP Asks You to Prove Your Publishing Rights that offered a few tips. It turns out this abuse of authors is totally normal and KDP staff is specifically instructed not to be helpful. Anyway, I sent them invoices from my editor and the full git history. I also included receipts from the above proof copies that I ordered.

The result? “We want … a shrubbery!”

I asked them again to clarify who the claimant is and which text it was about, but I received no reply. A week or so later I received their final (insulting) email:

The timestamp they refer to was one of the automated and useless emails I cited above.

And yes, I even tried emailing the infamous jeff@amazon.com address, with zero effect.

But Why?

Since Amazon has the policy of not telling authors why they reject your book, all I can do is guess. On the bright side, this lets me spout the most absurd conspiracy theories possible, perhaps boosting book sales in the process. Even though perhaps it was just a missing comma somewhere on the copyright page.

My best guess, though I think it’s unlikely, is that the problem is Appendix C, which is a copy of the Bitcoin Whitepaper. Since the book talks a lot about the design philosophy of Bitcoin, it seemed appropriate for me to include this. It also gave me a chance to improve the layout: everything looks better with a touch of Latex.

The whitepaper was included with the rest of the source code, which was released with the MIT license, see e.g. this archive (update 2022-06-06: I should have linked to this instead). The license is pretty clear:

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the “Software”), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software …

I proactively pointed this out in my emails to KDP “support”. I also pointed out that there is a well known imposter out there who falsely claims to own the copyright and who has been legally harassing those who publish the website. He most famously got bitcoin.org to block UK visitors from downloading, by taking advantage of the fact that the person behind that site needs to stay anonymous for his or her physical safety. Fortunately COPA is working on this pestilence.

Obviously I received no response on this.

Without knowing what the alleged “infringement claim” is about and who it’s from, I can’t take legal action against any false claims. In this manner Amazon is actively helping patent trolls and their kind. Rather than getting rid of spurious copyright claims, Amazon is forcing authors to seek the blessing of any potential copyright holder out there, even fraudulent ones.

The future

The most pernicious aspects of this censorship are that Amazon won’t let you speak to a human about it, and that they won’t tell you what the offence is. This is the most effective way to achieve self-censorship.

When your book is rejected, you won’t know if it’s a copyright issue, or if you’re just dealing with Amazon employee who doesn’t like Bitcoin, or if it’s because you said something bad about the CCP. And once you trigger the algorithm, you can be sure there will be additional scrutiny on you. The only way then to get your book published safely, is to stay as far away from anything that could trigger the algorithm. It won’t just be about phrasing things diplomatically – algorithms don’t care – you have to avoid certain topics completely: don’t talk about X directly, don’t mention X in passing, don’t joke about X, don’t use X in a metaphor.

Thankfully this future has not fully arrived yet. There are actually multiple companies that can print and distribute books. Not all of them apply the same opaque censorship AI as Amazon. But this seems merely a matter of time.

So Where’s The Book?

Ironically it’s available on Amazon¹I should say Amazons, because the many countries where they operate are often siloes. This can lead to a situation where Canadians get the book shipped from the UK, while Mexicans and Taiwanese get it from the US. , amongst other places.

I ended up doing my international distribution through IngramSpark. For the Dutch and Belgium market I use CB Compact, which runs the central book warehouse where all Dutch physical and online stores pull books from. They offer print-on-demand as well a hybrid form where they also keep inventory for faster shipping. In fact, two different German Amazon warehouses ordered a bunch of books from The Netherlands too. Who knows where the AI shipped those.

In any case, if you’re thinking about self publishing a paper book, consider using KDP for their quick and convenient proof copy service. Consider bringing your own ISBN rather than use the one Amazon gives you for free, so you’re not dependent on their wrath. Of course, many authors have had great success with them. Just know that there are other options too. For now.

Speaking of Amazon – Kindle Royalties Are Trash

Since the e-book version is out (which you can buy directly from me, pay with Bitcoin and even use Tor) I might as well continue my rant…

As a reader I love Kindle. I bought my first one in 2009. It came with free international 4G – back when mobile hotspots were less trivial – so I could impulse-buy a new book while riding a camel in Egypt.

As an author however I’m very disappointed by Kindle.

First of all epub is terrible format. Compared to a Latex generated PDF, it looks atrocious. It’s really just a fancy mobile website, and in fact it uses HTML and CSS. I don’t even want to know if you can run (malicious) javascript on certain e-readers. To be fair, this is not Kindle’s fault. Even Apple hasn’t bothered to fix this problem for their book store. How hard can it be to have an e-book standard based on Latex?

Second, let’s talk about royalties. Unless you sell your book for $9.99 or less, you get a mere 35%! When Bezos was asked about this in 2009 he basically gave the interviewer the middle finger:

In case it’s not obvious how Bezos is gaslighting the reader: the interviewer presumably had an old-school publisher, i.e. a company that pays you an advance, pays for an editor, takes physical inventory risk, etc. Some publishers may do a poor job at that – self publishing is popular for a reason – but Amazon does none of that.

Amazon charges a 186% fee²As a self-publisher, you are in the business of selling royalties. For every $3.50 worth of royalties that you earn from Amazon, they charge $6.50 extra to the customer. With less market power, they might charge $4.00 and still make a profit. for which their AI overloard may or may not get your book in front of more eyeballs than you can through your own marketing. They do handle payment processing, taxes and customer support, which is certainly worth something.

But you can buy the ebook elsewhere and put it on your Kindle? Sure, but’s more friction and it won’t render as nicely as when you get them through the store. E.g. it won’t show the cover image in your library. The Kindle may not be a fully walled garden, but there’s definitely a little fence.