The payout stage of Anthropic’s $1.5 billion copyright settlement has turned messy. Authors expecting cheques from the landmark deal say they instead received emails telling them somebody else had staked a claim on their money — in many cases their publisher, and in some cases their literary agent.
What the settlement actually says
Anthropic settled a copyright class action after a judge drew a sharp line: training AI models on copyrighted material can qualify as fair use, but pirating that material to do so does not. Final approval came through in July, clearing the way for payments.
The terms are simple on paper. Authors of close to 500,000 titles are due $3,000 per pirated work. If a book is still in print with a traditional publisher, author and publisher split the money 50-50. If the book was self-published, or the rights reverted because it went out of print, the author is meant to collect the full amount.
Where it is going wrong
Writers have taken to social platforms with complaints that publishers are claiming more than they are entitled to. Mystery and thriller novelist April Henry posted that HarperCollins had claimed one of her books whose rights reverted at least 17 years ago — and that on the same day she got a credit alert listing the publisher as her employer, which she says it never was.
Victoria Strauss, who runs the long-running Writer Beware blog, says the complaints reaching her fall into two buckets:
- Publishers claiming works whose rights have already reverted to the author
- Publishers claiming 100% of a payment when they are only entitled to 50%
- Literary agencies filing claims at all — which Strauss flags as odd, since agents are not rightsholders in the books they sell
Strauss said she is “reluctant to attribute to malice what can be plausibly explained by poor recordkeeping,” noting that some publishers have already called it an error and asked Anthropic to fix it. But she also wrote that the volume of reports, and the fact that authors are describing identical errors repeatedly, points to something “widespread and systemic” rather than routine glitches — while conceding her view is only “a peek through a small crack in a massive wall.”
Authors Guild CEO Mary Rasenberger struck a similar note, telling The New York Times she does not see this as “a grab by the publishers,” and blaming muddled records and a confusing claims process instead. Author Courtney Milan was less diplomatic about agents claiming percentages, posting on Bluesky that they should “stop that shit!”
Why marketers and media professionals should care
This is the first mass-scale test of what happens after an AI copyright settlement — and the lesson is that the legal win is the easy part. Distributing money at scale depends entirely on rights metadata that, in publishing, is often decades old, incomplete or contested.
Any content business sitting on licensable archives — publishers, broadcasters, stock libraries, agencies with client-owned creative — should read this as a warning. When AI licensing deals arrive, the party with clean, current, machine-readable rights records gets paid correctly. Everyone else spends months arguing.
Practical takeaways: audit reversion dates and rights ownership now, not when a cheque is announced; make sure contributor and freelancer agreements state who owns downstream licensing revenue; and build a single source of truth for rights rather than scattering it across contracts.
One detail matters for authors filing disputes: to claim 100% of a payment, the rights reversion must have happened before August 10, 2022, the settlement’s stated download date. Both Milan and the Authors Guild have shared guidance on how to challenge allocations.
Source: TechCrunch
