Photo was created by Webthat using MidJourney
Sarah Silverman and Authors File Lawsuits Against OpenAI and Meta Over Copyright Claims
Comedian Sarah Silverman, along with authors Christopher Golden and Richard Kadrey, has initiated legal proceedings against OpenAI and Meta, alleging copyright infringement. The lawsuits assert that the AI models developed by both companies, including OpenAI’s ChatGPT and Meta’s LLaMA, employed their copyrighted works without obtaining proper authorization.
ChatGPT and LLaMA: AI Models Under Scrutiny for Alleged Copyright Infringement
OpenAI’s ChatGPT and Meta’s LLaMA, widely known and utilized language models, are at the center of controversy as Silverman and the authors argue that their creative works were incorporated into the models’ training data without their consent. These models rely on large amounts of internet-derived information to generate coherent responses to user input.
Lawsuits Allege Unauthorized Use of Authors’ Works in AI Model Training
The lawsuit filed against OpenAI contends that Silverman, Golden, and Kadrey did not provide consent for their copyrighted books to be used as training material for ChatGPT. Nevertheless, their works were allegedly ingested and utilized during the training process. Similarly, the lawsuit against Meta claims that a significant number of the authors’ copyrighted books were included in the dataset employed to train LLaMA, an assortment of Meta-owned AI models.f
Acquisition of Works from “Shadow Library” Sites Raises Concerns
According to the lawsuits, the authors’ works were sourced from “shadow library” sites, which have long piqued the interest of the AI training community. OpenAI’s lawsuit includes exhibits demonstrating ChatGPT’s ability to summarize three specific books: Silverman’s “The Bedwetter,” Golden’s “Ararat,” and Kadrey’s “Sandman Slim.” The Meta suit highlights multiple works by Kadrey and Golden, along with “The Bedwetter,” and raises the issue of Meta’s paper that acknowledges LLaMA’s training datasets containing material from these shadow libraries, deemed “flagrantly illegal.”
Legal Representatives Voice Concerns Regarding AI Models’ Text Generation
Attorneys Joseph Saveri and Matthew Butterick, representing the three authors, have reported receiving concerns from writers, authors, and publishers regarding ChatGPT’s remarkable ability to generate text that closely resembles copyrighted material. Saveri and Butterick are also representing Mona Awad and Paul Tremblay in a separate class action lawsuit against OpenAI, alleging unauthorized use of their works during ChatGPT’s training. Additionally, the legal team is involved in lawsuits on behalf of artists Sarah Andersen, Kelly McKernan, and Karla Ortiz against image generators Stability AI, Midjourney, and DeviantArt.
Lawsuits Highlight AI Models’ Potential for False Answers and Defamation
Apart from copyright infringement, the lawsuits raise concerns about the occurrence of false answers, often referred to as “hallucinations,” produced by AI models. In a separate case, a radio host from Georgia has sued OpenAI for defamation after ChatGPT falsely implicated him in fraud. These legal actions shed light on the need for responsible AI development and its impact on various industries and individuals.