Harry Potter publisher to receive millions in Anthropic copyright settlement

Bloomsbury has 14,087 titles listed in agreement between AI startup and authors over use of their work

Harry Potter publisher to receive millions in Anthropic copyright settlement

Bloomsbury, the publisher famous for the Harry Potter books, will receive millions of pounds as part of a large settlement. This agreement is between the artificial intelligence company Anthropic and many authors whose work was used to train AI chatbots.

Bloomsbury, which publishes well-known authors like Sarah J Maas and Susanna Clarke, as well as J.K. Rowling, stated that 14,087 of its titles were included in this settlement. Each title is expected to receive approximately $3,000 in compensation.

The London-based company anticipates receiving this money in stages, possibly beginning in the latter half of the current financial year. The funds will be shared between Bloomsbury and the authors whose works were used.

After a deduction of about 10% to cover legal fees and other costs, Bloomsbury and the authors involved are expected to receive roughly $19m. This situation highlights a significant issue in the rapidly growing AI industry.

Copyright has become a major point of contention in the AI boom. AI tools, such as Anthropic's Claude chatbots, are developed using huge amounts of data collected from the internet. This data often includes copyrighted material like novels and news articles.

AI companies in the US have argued that using such material falls under the legal principle of fair use, which permits the use of copyrighted work without permission in certain situations. However, artists and creative professionals globally have been demanding that AI companies obtain permission first, or at least compensate them for using their creations.

A US district judge, Araceli Martínez-Olguín, stated that the settlement offers significant support to the authors and publishers affected. The legal action was initiated in 2024 by novelist Andrea Bartz and two other authors. Approximately 91% of the 482,000 works covered by the lawsuit have had claims submitted.

The lead lawyer for the authors, Justin Nelson, described the settlement as the largest known recovery for copyright issues in history. This is the first major resolution among many AI copyright lawsuits currently being processed in US courts on behalf of authors and news organisations.

Anthropic’s deputy general counsel, Aparna Sridhar, expressed satisfaction that a large majority of authors and publishers involved have claimed their payment. She noted that the company looks forward to finalising the matter. Last year, Bloomsbury also established an AI licensing agreement to permit the use of academic works for training AI programmes.


Vocabulary

settlement — a formal agreement to end a dispute, often involving payment
AI startup — a newly formed company focused on artificial intelligence technology
compensation — money paid to someone for loss or injury, or for work done
instalments — payments made over a period of time, rather than all at once
battleground — a subject or area where there is strong disagreement or conflict
trained on — taught or developed using a large amount of data
fair use — a legal principle allowing limited use of copyrighted material without permission for purposes like criticism or education
doctrine — a belief or set of beliefs held and taught by a church, political party, or other group
generative AI programmes — AI systems capable of creating new content, such as text or images
royalties — a payment made to an author or creator based on the sales or use of their work

Discussion Questions

  1. Why has copyright become such an important issue in the development of AI?
  2. What is the main argument used by AI companies to justify using copyrighted material without permission?
  3. How might this settlement impact future agreements between AI companies and creators?

Based on an article from The Guardian.

Read the original article