By; OLIVIA NNOROM
Comedian Sarah Silverman, Christopher Golden and Richard Kadrey sue Meta Platforms and OpenAI, alleging copyright infringement in training AI language models.
Plaintiffs claim Meta and OpenAI used their books without permission to develop large language models.
Lawsuits highlight legal risks faced by developers using copyrighted material to create chat bots with realistic responses.
Sarah Silverman, US comedian and two other authors have filed copyright infringement lawsuits against facebook parent platform, Meta and ChatGPT developer, OpenAI, alleging that their copyrighted content was used without permission to train artificial intelligence language models.
The proposed class action lawsuits, filed in San Francisco federal court, claim that Meta and OpenAI utilised the plaintiffs’ books to develop their large language models, designed to replicate human conversation.
In the OpenAI lawsuit, ChatGPT is accused of summarising three books: “The Bedwetter” by Sarah Silverman, “Ararat” by Christopher Golden, and “Sandman Slim” by Richard Kadrey. The Meta lawsuit mentions multiple works by Golden and Kadrey, in addition to Silverman’s “Bedwetter.”
Furthermore, the lawsuit highlighted a leaked information about Meta paper that suggested LLaMA’s training datasets included materials from Shadow libraries, describing this action as “flagrantly illegal.”
“The summaries get some details wrong” but still show that ChatGPT “retains knowledge of particular works in the training dataset,” the lawsuit says.
Beyond the lawsuits, Joseph Saveri and Mattew Butterick, the lawyers representing the plaintiffs have claimed that ChatGPT have attracted complaint from writer, publishers and authors who pointed out the ability of the AI tool to generate articles that look like plagiarised materials
Meta and OpenAI, a company backed by Microsoft, have not responded to the allegations. These lawsuits highlight the legal risks associated with using copyrighted material to train AI models and develop chatbots that deliver realistic responses to user prompts.
The plaintiffs are seeking unspecified monetary damages on behalf of a nationwide class of copyright owners whose works were allegedly infringed upon.
The outcome of these lawsuits could have implications for developers utilising copyrighted material to train AI models and prompt discussions about the boundaries of fair use in the context of machine learning and natural language processing.
Comments 2