Skip to content

Court Bolsters Authors' Lawsuit Against OpenAI, Advances New Theories of Infringement

The court's ruling could force OpenAI to pay significant damages. The case may set a precedent for AI's use of copyrighted material.

This is book.
This is book.

Court Bolsters Authors' Lawsuit Against OpenAI, Advances New Theories of Infringement

A federal court has advanced two new theories of infringement against OpenAI, bolstering a lawsuit brought by George R.R. Martin and other Authors Guild members. The authors accuse OpenAI of copyright infringement and theft by training its language model, ChatGPT, on pirated works of fiction.

The court ruled that a summary of Martin's 'A Song Of Ice And Fire' generated by ChatGPT is substantially similar to the original novel, supporting the first theory of infringement. This theory suggests that training language models on copyrighted books without authors' consent constitutes infringement. OpenAI argued against separating the two theories, but the court sided with the authors.

The second theory alleges that OpenAI's practice of downloading titles from 'shadow libraries' full of stolen books constitutes piracy. The court determined that outputs about potential or alternative sequels to Martin's books using his original characters could also be seen as infringing on his works. A third theory suggests that ChatGPT's answers can infringe on the books they're trained on if they are substantially similar to the original works.

The authors are building a strong case, which may force OpenAI to pay significant damages before the trial. The court only needs to rule in favor of the authors in one of the theories to impose statutory damages on OpenAI.

OpenAI faces a significant legal challenge with these new theories of infringement. The company may have to pay substantial damages if the court rules in favor of the authors. Meanwhile, OpenAI has not publicly defended its use of ChatGPT in light of these allegations, and the case continues to unfold.

Read also:

Latest