Publishers Demanded Answers From Nippan on Anthropic Sales
The Japan Book Publishers Association has inquired about potential copyright infringements tied to AI model training.
Updated on Sept. 21, 2026 in Artificial Intelligence
Live Poll
Should AI companies be required to obtain permission before using copyrighted books for training?
The Japan Book Publishers Association, representing 380 companies, has requested an explanation from distributor Nippan Group Holdings regarding alleged book sales to AI developer Anthropic. While Nippan has denied selling books for AI training purposes, internal Anthropic documents referenced in U.S. court proceedings claim a sales contract was established in November 2024.
Why it matters
The dispute centers on growing concerns among publishers regarding unauthorized book usage for large language model development. It highlights the friction between AI firms seeking high-quality training datasets and copyright holders aiming to protect intellectual property.
The inquiry follows claims of a sales contract signed in November 2024, which surfaced in U.S. litigation involving Anthropic. These documents allege the use of a website for pirated copies for model learning purposes.
The players
Japan Book Publishers Association
A trade organization representing 380 Japanese publishing companies concerned with copyright enforcement and intellectual property rights.
Nippan Group Holdings
A major Japanese book distributor currently facing scrutiny over its reported dealings with artificial intelligence firms.
Anthropic
An AI research and deployment company that develops large language models and has recently faced copyright litigation in the United States.
The details
The association is seeking clarification on the nature of these transactions and the specific legal justifications used by Nippan. Anthropic, a company focused on developing large language models—complex AI systems trained on vast datasets to predict and generate text—has faced previous copyright challenges. This includes a settlement reached with American writers in late July 2026, where U.S. court documents first drew connections between the developer and Nippan.
Timeline
November 2024: Anthropic allegedly signed a sales contract with Nippan.
Late July 2026: Anthropic settled a copyright lawsuit with American writers.
September 21, 2026: The inquiry regarding the alleged sales was reported.
The Tech Race
This inquiry follows the legal precedents set by the Anthropic settlement with American writers in July 2026. It signals a move toward global accountability for the datasets used to train competitive artificial intelligence models.
The outcome of this inquiry will likely influence how international distributors handle data requests from AI developers. Publishers and content creators should monitor these proceedings to understand if their distribution contracts will include new safeguards against AI-driven data ingestion.
The takeaway
This case underscores the widening legal net cast over the data sourcing practices of generative AI companies. Stakeholders should track future disclosures from Nippan and any subsequent legal filings by the Japan Book Publishers Association to see if a formal challenge follows the inquiry.
Further reading
For broader context on the ongoing legal and ethical challenges in machine learning, visit Artificial Intelligence.
Source note: This article includes information reported by 朝日新聞デジタル.
Live Poll
Should AI companies be required to obtain permission before using copyrighted books for training?
