Writer and performer Sarah Silverman is suing Chat GPT creator OpenAI and Facebook owner Meta for copyright infringement at a California district court along with two other writers. 

OpenAI created and maintains ChatGPT, which works by a large language model being trained by copying massive amounts of text and extracting ‘expressive’ information from it. A large language model’s output is entirely and uniquely reliant on the material in its training dataset.

In the lawsuit, the plaintiffs, who also include writers’ Christopher Golden and Richard Kadrey, allege that much of the material in OpenAI’s training datasets comes from copyrighted works “including books written by plaintiffs, that were copied by OpenAI without consent, without credit and without compensation.”

Their permission was not sought for their copyrighted materials to be “ingested” and used to train ChatGPT.

In the Meta lawsuit, the plaintiffs again claim that Meta’s AI product LLaMA, which works like ChatGPT being a set of large language models, come from copyrighted works including books written by the plaintiffs “that were copied by Meta without consent, without credit and without compensation.”

The plaintiffs in both suits are represented by the law firm Joseph Saveri and Los Angeles lawyer Matthew Butterick, a specialist in AI, copyright and software law. 

This is one of a number of lawsuits to be filed against OpenAI: on 28 June a class action lawsuit was lodged at the same Californian court by 16 individuals accusing, among it other things, of violating the Electronic Communications Privacy Act and California’s Unfair Competition Law Business and Professions Code.

In this separate lawsuit the plaintiffs point out that the defendant OpenAI was originally founded as a nonprofit research organisation with a single mission: to create and ensure artificial intelligence would be used for the benefit of humanity. But in 2019, OpenAI abruptly restructured itself, developing a for-profit business that would “pursue commercial opportunities of staggering scale”.

As a result of the restructuring, OpenAI abandoned its original goals and principles, “electing instead to pursue profit at the expense of privacy, security, and ethics. It doubled down on a strategy to secretly harvest massive amounts of personal data from the internet.” 

It accuses ChatGPT of stealing private information, “including personally identifiable information, from hundreds of millions of internet users, including children of all ages, without their informed consent or knowledge.”

It continues that defendants’ disregard for privacy laws is matched only by their “disregard for the potentially catastrophic risk to humanity.” 

Chair of the intellectual property department at Cole Schotz, William Stroever, said that the  lawsuit filed by Sarah Silverman is unfortunately not a “surprising one, and it mirrors some of the lawsuits already filed by artists against AI projects like Midjourney and Stability AI”. 

He explained that the artificial intelligence typically generates outputs by studying existing works in training datasets and then creating outputs in response to a prompt. The question of whether the actual AI process is an infringement or whether an output is an infringement of an existing work is a "difficult question to answer".

He continued  that on  the other hand, human action in copying copyrighted works for inclusion into training data sets is much more familiar to the existing legal landscape. 

“These are probably the types of copyright cases we are likely to continue to see first as courts start to wade into the artificial intelligence landscape. As we move further forward, we are likely to then start seeing more ownership-related litigation, although some of that has started already as the US Copyright Office and  the US Patent and Trademark Office struggle with similar questions," he concluded.