Sarah Silverman files civil lawsuit against OpenAI and ChatGPT, claiming AI learning infringed her copyright

Last week we reported that OpenAI’s popular ChatGPT artificial intelligence chatbot was being sued for copyright infringement by two authors. And over the weekend, Sarah Silverman has become the latest personality to take aim at both ChatGPT and Meta, the owners of Facebook, Instagram and current threat to Twitter, Threads.

The Daily Beast reported that Richard Kadrey and Christopher Golden have joined Silverman in the suits, but according to the lawsuit, the number of potential plaintiffs is “at least thousands of members.” Much akin to the lawsuit filed by Paul Tremblay and Mona Awad, the plaintiffs cite the use of “shadow libraries” containing unlawful copies of literature penned by those pursuing the case.

In Silverman’s case, she is alleging her 2010 memoir, “The Bedwetter,” was used without her permission as part of ChatGPT’s “mechanised learning,” and therefore was in breach of her copyright using the material without her permission. 

With regards to the filing against Mark Zuckerberg’s Meta company, all three plaintiffs have claimed that “many” of their books appear in the dataset that the Facebook and Instagram owner used to train LLaMA, a group of Meta-owned AI models. “If a user prompts ChatGPT to summarise a copyrighted book, it will do so,” the suit against OpenAI claims.

Both suits claim these language models are illegal under the copyright, since they need copyrighted materials in order to “write” work of their own

Under US copyright law, AI programs might infringe copyright by generating outputs that resemble existing works; case law, copyright owners may be able to show that such outputs infringe their copyrights if the AI program both (1) had access to their works and (2) created “substantially similar” outputs. 

This is where the use of “shadow libraries,” cited as a popular reference point for AI language models to learn from, might be the important aspect of the cases OpenAI and Meta are now having to contend with.