Authors dispute publishers' claims on Anthropic copyright settlement payments
Authors reported that publishers were claiming excessive shares or rights to Anthropic copyright settlement payments to which the authors were entitled. Industry figures attributed the disputes to poor recordkeeping and settlement administration complexity rather than deliberate misconduct by publishers.
What happened
Anthropic settled a copyright class action suit, with a judge ruling that training AI models on copyrighted material is legal under fair use doctrine, but pirating that material was not. The $1.5 billion settlement received final approval in July 2026, with the authors of nearly 500,000 titles to be paid $3,000 for each pirated work.
Under settlement terms, if a book is still in print with a traditional publisher, the money will be split 50-50 between author and publisher. If the book was self-published or the publisher reverted the rights, the author should get the entire payment. Authors reported receiving emails informing them that publishers were making claims on their payments.
Reported impact
Authors posted on social media that publishers were claiming more than their fair share. Mystery and thriller author April Henry reported that HarperCollins claimed one of her books that had reverted rights at least 17 years earlier, while simultaneously adding her as an employee in error.
Victoria Strauss reported receiving author complaints in two categories: publishers seeking payment for works they no longer have legitimate claim on due to reverted rights, and publishers seeking full 100% payment when entitled only to 50%.
Index analysis
Victoria Strauss noted reluctance to attribute the claims to malice, explaining that poor recordkeeping could account for the errors, and that some publishers stated these were mistakes they had asked Anthropic to fix.
Authors Guild CEO Mary Rasenberger stated she did not view this as a grab by publishers and did not believe they were specifically trying to harm authors, instead characterizing it as a result of bad recordkeeping and a confusing settlement process. Strauss acknowledged the complaints represented only initial reports, but noted that the unusually large number of reports and authors reporting identical errors suggested these were not routine glitches.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.