Copyright, Publishers, and Your Anthropic Payout
In October, a settlement was announced in the copyright lawsuit against Anthropic, providing authors of books they trained their AI on with compensation of around $3000. There were certain eligibility conditions, though.

One condition is that the book must have been registered with the US Copyright Office during the period in which Anthropic downloaded it. You can check whether yours is registered here. Traditionally, and usually by contract, publishers are responsible for registering books with the US Copyright Office.
At least one author has learned that her publisher failed to register her book, even though it was published several years ago. So it seems her publisher’s dawdling has cost her around $3000.
The publisher in question is Routledge, of the Taylor & Francis Group. The author is philosophy professor Sarah Paul (NYU Abu Dhabi). She writes:
I checked whether my book (which had been published before Anthropic’s downloading ended, and is on LibGen [a site scraped by Anthropic]) was registered, and it still isn’t. This is now 5 years after my book was published. I have no way of knowing whether Anthropic actually downloaded my book, but it is possible that my book was excluded from the settlement because the copyright was not registered. Further, my contract with T&F states that they must register it:
“The Publishers will register the copyright in the Work in the name of the Publishers in compliance with the copyright law of the United States, and will publish the Work with a copyright notice in the name of the Publishers.”
I raised my concern with T&F that their failure to register the copyright could have cost me ~$3k. I believe some other presses are actively seeking information for their authors about whether a failure to register the copyright was the reason for exclusion, and if so, are compensating their authors accordingly. I suggested that this is something T&F could also do. I waited 8 weeks for an official response, but eventually got nothing other than an offer to register my copyright now, “if I would like.”
This is infuriatingly irresponsible behavior from Taylor & Francis, capped by an infuriating irresponsible reply. Authors may wish to keep this in mind next time they’re considering publishing a book with Routledge or any of T&F’s other imprints.
In the meanwhile, Professor Paul is curious whether other authors are in a similar position of having been excluded from the settlement owing to publishers failing to register the copyright of their works (be it Taylor & Francis or another publisher). If so, please let us know in the comments. She adds, “Perhaps some action could be taken together, if there are enough of us.”
Neither my Routledge monograph nor my edited volume for them are registered. But this also seems to be the case with pretty much all of my other books on LibGen (whose publishers include CUP and Palgrave) with the exception of Bloomsbury. Wiley-Blackwell have registered an anthology I co-edited with Jonathan Dancy, but not the Companion to the Philosophy of Action I co-edited with Tim O’Connor (both fall within the eligible publication period).
My Cambridge monograph is on LibGen too and it does not seem to have been registered. I had no idea about any of this! When news of the settlement came out, I did see my book in the list, but I thought for some reason they had evidence it was not used. I did not know LibGen was scraped (my book was there basically from day one).
Thanks for posting this. I see neither of my two books (one publilshed by Continuum (now Bloomsbury) and the other by Oxford University Press) is registered with the US Copyright Office. I fear their oversight has cost me a good bit of money (more than I will make from royalties!)
Sarah, I would be interested in a collective action remedy.
Only one of five books published under my name have been registered. The ones not registered are from OUP UK, CUP, Routledge, and Rodopi (now something else). The one that was registered was published with OUP USA. So I wonder whether non-US presses typically do not register copyright in the US and if so, whether there would be any other kind of remedy for people who published with non-US presses.
On a separate matter, I don’t think we’ll make anywhere near $3000 per book. The notice I got said that it is $3000 “prior to the deduction of any costs, fees, and expenses”. If these include lawyers’ fees (it was not clear from the email whether these are awarded separately), they will eat up quite a chunk of the award. Moreover when I filled the form of the claim, the “default” division of the claim seemed to be 50/50 between “beneficial” owner (author) and “legal” owner (publisher). You don’t need to go for the default, but deviating seemed to trigger a mountain of forms.
Please note that the last point is under the caveat that I spent the minimum amount of time reading the relevant documents compatible with my signing the “yeah, i understand it” clause in reasonably good conscience.
Yes, I’d be surprised if we made $300
Fair enough — the charge of $3k in damages is probably greatly exaggerated, though a good starting point. The default 50/50 split with the publisher doesn’t apply to textbooks (which my book is), but I doubt this would end up working in my favor; it may be that the publisher thinks they would be entitled to the whole sum anyhow.
I have 5 titles in libgen with Packt Publishing and none seem to be registered. I didn’t know this was a thing. That’s pretty infuriating.
The other evening, I received a detailed and explanatory email from the Director of University of Michigan Press which has published three of my books: one monograph and two editions of an interdisciplinary collection. They recommended that I file a claim for one of my books because it had been used by Anthropic. The letter also indicated that the $3000 would be split evenly between the legal team, U of M Press, and me.