Chat GPT and Hollywood
Chat GPT logo Image Credit: Reuters

Some of the world's best-known novelists banded together this week to sue ChatGPT maker OpenAI for using their work to train its artificial intelligence tools, adding their efforts to the growing group of artists, musicians and writers trying to stop tech companies from benefiting from their work without paying for it.

Blockbuster writers such as George R.R. Martin, Jodi Picoult, Jonathan Franzen and George Saunders have signed on to the lawsuit, which is being led by the Authors Guild, a group that lobbies on behalf of writers.

It adds to a growing list of copyright lawsuits against OpenAI and other AI companies, including one from comedian Sarah Silverman and another from novelists Mona Awad and Paul Tremblay.

This latest lawsuit alleges OpenAI copied the authors' work "wholesale, without permission or consideration" and used it to train their "large language models" - the giant algorithms that power tools like ChatGPT.

"At the heart of these algorithms is systematic theft on a mass scale," the lawsuit says.

The lawsuit is the latest salvo in the ongoing debate over how AI tools should be trained and whether the companies behind them owe anything to the original creators of the training data. Large language models are generally trained on billions of sentences of text pulled from the internet, including news stories, Wikipedia and comments on social media sites. OpenAI and other AI companies such as Google and Microsoft do not say specifically what data they use, but AI critics have long suspected that it includes well-known collections of pirated books that have circulated online for years.

In the lawsuit, the authors ask for damages for the "lost opportunity to license their works" and for an injunction against OpenAI to bar it from continuing to use their work in its training data.

A spokesperson for OpenAI declined to comment on the pending litigation.

The tech companies have argued their use of data scraped from the internet to train AI is legal under the concept of fair use - a provision in copyright law that allows people to draw on the work of others if the final output is sufficiently different from the original. Content owners have pointed to the fact that AI tools often create images that are extremely similar to original work from humans as evidence that their creations are being copied.

Striking Hollywood writers and actors have said they want commitments from TV and movie production companies that AI won't be used to supplant their work. News organizations have blocked AI companies from scraping their websites. At the same time, some publishers have signed deals to sell their content directly to the tech companies. The Associated Press licensed its archive to OpenAI, and Universal Music Group struck a deal with YouTube to run experiments on how AI could be used to create music.