Comedian and author Sarah Silverman, as well as authors Christopher Golden and Richard Kadrey \u2014 are suing OpenAI<\/a> and Meta<\/a> each in a US District Court over dual claims of copyright infringement<\/a>. <\/p>\n<\/div>\n The suits alleges, among other things, that OpenAI\u2019s ChatGPT and Meta\u2019s LLaMA were trained on illegally-acquired datasets containing their works, which they say were acquired from \u201cshadow library\u201d websites like Bibliotik, Library Genesis, Z-Library, and others, noting the books are \u201cavailable in bulk via torrent systems.\u201d<\/p>\n<\/div>\n Golden and Kadrey each declined to comment on the lawsuit, while Silverman\u2019s team did not respond by press time.<\/p>\n<\/div>\n In the OpenAI suit, the trio offers exhibits<\/a> showing that when prompted, ChatGPT will summarize their books, infringing on their copyrights. Silverman\u2019s Bedwetter <\/em>is the first book shown being summarized by ChatGPT in the exhibits, while Golden\u2019s book Ararat <\/em>is also used as an example, as is Kadrey\u2019s book Sandman Slim<\/em>. The claim says the chatbot never bothered to \u201creproduce any of the copyright management information Plaintiffs included with their published works.\u201d<\/p>\n<\/div>\n As for the separate lawsuit against Meta, it alleges the authors\u2019 books were accessible in datasets Meta used<\/a> to train its LLaMA models, a quartet of open-source AI Models<\/a> the company introduced in February. <\/p>\n<\/div>\n The complaint lays out in steps why the plaintiffs believe the datasets have illicit origins \u2014 in a Meta paper detailing LLaMA<\/a>, the company points to sources for its training datasets, one of which is called ThePile, which was assembled by a company called EleutherAI. ThePile, the complaint points out, was described in an EleutherAI paper<\/a> as being put together from \u201ca copy of the contents of the Bibliotik private tracker.\u201d Bibliotik and the other \u201cshadow libraries\u201d listed, says the lawsuit, are \u201cflagrantly illegal.\u201d<\/p>\n<\/div>\n In both claims, the authors say that they \u201cdid not consent to the use of their copyrighted books as training material\u201d for the companies\u2019 AI models. Their lawsuits each contain six counts of various types of copyright violations, negligence, unjust enrichment, and unfair competition. The authors are looking for statutory damages, restitution of profits, and more.<\/p>\n<\/div>\n Lawyers Joseph Saveri and Matthew Butterick, who are representing the three authors, write on their LLMlitigation website<\/a> that they\u2019ve heard from \u201cwriters, authors, and publishers who are con\u00adcerned about [ChatGPT\u2019s] uncanny abil\u00adity to gen\u00ader\u00adate text sim\u00adi\u00adlar to that found in copy\u00adrighted tex\u00adtual mate\u00adri\u00adals, includ\u00ading thou\u00adsands of books.\u201d<\/p>\n<\/div>\n Saveri has also started litigation against AI companies on behalf of programmers<\/a> and artists<\/a>. Getty Images also filed an AI lawsuit<\/a>, alleging that Stability AI, who created the AI image generation tool Stable Diffusion, trained its model on \u201cmillions of images protected by copyright.\u201d Saveri and Butterick are also representing authors Mona Awad and Paul Tremblay in a similar case<\/a> over the company\u2019s chatbot. <\/p>\n<\/div>\n Lawsuits like this aren\u2019t just a headache for OpenAI and other AI companies; they are challenging the very limits of copyright<\/a>. There\u2019s As we\u2019ve said on The Vergecast<\/em> every time someone gets Nilay going on copyright law, we\u2019re going to see lawsuits centered around this stuff for years to come<\/a>.<\/p>\n<\/div>\n