This image shows a screen displaying the OpenAI and ChatGPT logos. AFP
Plaintiffs accuse the San Francisco company of using its copyrighted work without permission to train artificial intelligence models, a series of actions that could complicate the development of one of tech’s biggest emerging trends. Incidents will increase.
The three also filed a lawsuit against Facebook’s parent company, Meta. Meta’s lesser-known open-source model also used pirated downloads of books for training purposes, he said.
Many of the training materials used by OpenAI and Meta are derived from “copyrighted works, including books written by Plaintiffs, which are not consented to or credited by OpenAI and Meta.” , has been copied without compensation,” the three lawyers said in a blog post.
In both cases, filed in California court on Friday, the authors accuse the tech company of using their books to train AI models and allege a series of copyright infringements.
Successful lawsuits of this sort would upend the way technology is developed, limiting the way tech giants build models and churning out compelling, human-like content.
Plaintiffs in a recent concentration action lawsuit include source code owners against OpenAI and Microsoft’s GitHub, visual artists, and photography agency Getty against Stability AI.
San Francisco attorneys Joseph Savelli and Matthew Batterrick, who have led other similar cases, have filed the latest lawsuit on behalf of Silberman and authors Christopher Golden and Richard Cudley.
The lawsuit cited Silverman’s 2010 best-selling memoir “The Bedwetter,” Golden’s horror novel “Ararat,” and Cudley’s “Sandman Slim” paranormal noir series.
Silberman is best known in the United States for his edgy and often controversial humor, as well as his outspokenness on social and political issues.
To OpenAI, the plaintiffs “did not agree to use their copyrighted books as training materials for ChatGPT. , was used to train ChatGPT.”
In their lawsuit, the authors provided evidence outlining ChatGPT’s work in detail.
The three allege to Meta that Meta used an illegally constructed “shadow library” to build its LLaMA model, including their own work.
These libraries use pirated torrent downloads to illegally publish copyrighted works.
OpenAI declined to comment on the lawsuit, but Meta did not immediately respond to a request for comment.
short link:
