AI Copyright Infringement Case to Commence

Sarah Silverman is set to sue the developer of Chat GPT, Open AI and Meta for copyright infringement over claims that these platforms were trained using work of hers and other authors such as Christopher Golden and Richard Kadrey.

Chat GPT and similar systems are built to learn to imitate human language. To do so, it must analyse large amounts of data, including human texts.

The trio claim that Open AI did not gain their consent before using copyrighted books and material for Chat GPT, yet their materials were used to train Chat GPT via “shadow library” sites which the claimants describe as “flagrantly illegal”.

Silverman is also pursuing a claim against Meta, (the owners of social media groups such as Instagram and Facebook), on the grounds that the LLaMA AI system, designed to help AI research, has been trained by a system which her books have appeared in a dataset for.

The claim suggests that works such as The Bedwetter, Ararat and Sandman Slim have all been used and the lawyers of the trio claim to have heard from many other authors, publishers and writers with concerns about generated texts and their similarity to copyrighted works.

This is not the only infringement issue that has been brought to light recently, with Getty Images suing Stability and other artists such as Kelly McKernan commencing lawsuits against AI image generators such as DeviantArt. Open AI are also facing lawsuits from Mona Awad and Paul Tremblay on similar grounds.

However, with Open AI claiming that Chat GPT is only trained with data that is readily available to the public through the internet, it’s unclear on how such lawsuits will develop and conclude. Despite this, Andres Guadamuz from the law department of the University of Sussex suggests this will be the first lawsuit against Chat GPT’s exploration of the uncertain “borders of the legality”.

With claims of unlawful absorbency of copyrighted works and “very accurate summaries” of such novels included in the exhibits of the lawsuits, much is to be deciphered. Patrick Goold from City University in London proclaims that he believes the success or otherwise of cases brought by Silverman will likely come down to whether training a large language model is “fair-use" or not, with other experts questioning whether Open AI can be said to have copied works.

The law firm that is assisting the writers have also launched similar cases last year on behalf of programmers and other artists who believed that a variety of AI systems had been infringing their rights as authors/creators. Getty Images says infringement by Stability AI’s Stable Diffusion system had been committed, causing a rise of worry on social media platforms including a tweet that claimed, “This thing wants our jobs- it’s actively anti-artist”.

Getty Images claims, “Stability AI unlawfully copied and processed millions of images protected by copyright... to the detriment of the content creators" and "chose to ignore viable licensing options and long‑standing legal protections". However, a Stability AI representative argued that such a claim “misunderstands the law” and that those who believe such claim is fair use, they do “not understand the technology”.

If someone has used your copyrighted work without permission, we will be able to help. Contact us to speak to one of our experts today. 

Article credit: Emily Clemence