Authors expecting payments from Anthropic’s $1.5 billion copyright settlement are encountering a new dispute: publishers and literary agencies are also claiming portions of the money. TechCrunch reports that some writers received emails this week saying another party had asserted a right to payments attached to their books.
The complaints do not yet establish deliberate misconduct or show how widespread the errors are. They do expose how a settlement covering nearly 500,000 titles depends on publishing records that may be old, confusing or incomplete. With thousands of dollars allocated to each work, uncertainty over who held which rights at a specific moment has immediate financial consequences.
A simple payment formula meets complicated rights histories
Anthropic settled the copyright class action last year after a judge distinguished between two parts of the company’s conduct. Training AI models on copyrighted material was deemed lawful under fair use, while pirating the material was not. Final approval arrived in July, allowing payments to proceed.
Under the settlement terms described by TechCrunch, each pirated work receives $3,000. For a traditionally published book that remains in print, the author and publisher split that amount equally. An author is entitled to the entire payment when a book was self-published or when a publisher allowed it to go out of print and returned the rights.
That framework appears clear until it is applied to decades of contracts and changing ownership. A book may have moved between in-print and out-of-print status, while rights could have reverted long before the settlement. The process also imposes a firm historical cutoff: an author seeking the full amount must show that reversion occurred before August 10, 2022, the settlement’s designated download date.
The resulting allocation exercise is therefore not merely a list of titles and names. It requires accurate evidence about publication status, contractual rights and timing for every disputed work.
Authors report recurring claims that do not fit the rules
Victoria Strauss of the Writers Beware blog said the complaints reaching her fall into two broad groups. In one, publishers claim a share for books whose rights had already reverted to the author. In the other, publishers seek the full payment even though the settlement would give them only half.
Some publishers have said claims were submitted by mistake and asked Anthropic to correct them. Strauss cautioned against assuming malicious intent when poor records could explain the problem. Authors Guild chief executive Mary Rasenberger offered a similar assessment, arguing that the reports look more like the predictable outcome of confusing procedures and bad recordkeeping than a coordinated effort to seize authors’ money.
Even so, Strauss said she had received an unusually high number of complaints over two days and that writers were repeatedly describing the same errors. In her view, that pattern may point to a broader systemic problem rather than the routine glitches expected in an operation of this size.
One author cited by TechCrunch said a publisher claimed a book even though its rights had reverted at least 17 years earlier. Such accounts remain individual reports, but they show why an incorrect rights record can redirect a meaningful payment.
Agency claims add another layer of uncertainty
Publishers are not the only parties appearing in the disputed allocations. Strauss said authors have also complained about literary agencies seeking percentages. She found those claims surprising because agents are not the rightsholders in the books they sell. Author Courtney Milan publicly objected to agencies taking part of the settlement, while she and the Authors Guild shared instructions for challenging allocations.
TechCrunch does not report how many publisher or agency claims have been challenged, how many have been corrected or what share of all eligible titles may be affected. Strauss described the complaints she had seen as only a narrow view into a much larger process. That limitation makes it too early to quantify the problem.
The dispute nevertheless marks a new phase in the Anthropic case. The central issue is no longer only whether the company must compensate authors for pirated books; the settlement has already answered that question and defined the amounts. Now the process must determine who receives each payment, using records capable of distinguishing current publishing relationships from rights that ended years ago.
For affected writers, the dispute procedure is the immediate route to correction. For the settlement as a whole, repeated claims on reverted books or incorrect percentages will test whether its administration can translate a legal formula into accurate payments across an unusually large and complicated catalog.
