Writers George RR Martin, John Grisham (The Firm, The Pelican Report), Michael Connelly (The black box) and Elin Hilderbrand, among others, through the Authors Gild union, have denounced the technology company OpenAI, creator of the chatbot with generative artificial intelligence (AI) ChatGPT, for violating their copyright and for a “systematic theft on a massive scale” of his work.
🎙 PODCAST | “Naked whoever you want”: artificial intelligence is getting out of hand
The writers pointed out to the Southern District of New York that the company – which gained widespread popularity last year with the launch of ChatGPT – incorporated the content of its books into its “language models” (LLM), which train its generative AI, capable of creating content in a sophisticated language in imitation of the human.
In the complaint, which was filed this Tuesday but became known on Wednesday, it states that ChatGPT users have used this tool to write prequels and sequels to the bestseller such as Song of ice and fire by Martin, a series of fantasy novels that was adapted into the hit HBO series Game of Thrones. Martin has not yet published the final two novels of this plot.
The authors want the court to prohibit OpenAI from using copyrighted works in their LLMs without “express permission” and are seeking damages including up to $150,000 per work.
This is not the only such complaint that OpenAI faces. This year, comedian Sarah Silverman and authors Christopher Golden and Richard Kadrey sued OpenAI and Meta for copyright infringement to train GPT-4 and Llama 2.
OpenIA is not the only technology company that has launched chatbots with generative AI; other companies, such as the titans Google, Meta and Microsoft, also offer tools of this type for free.
Until now, technology companies defend that their chatbots do not plagiarize the content of artists, but rather are inspired by their works to create their own content.
In August, OpenAI said that website owners can now block their web crawler, from allowing their content to be used to train their LLM. Several sites, including news publishers such as The New York Times, have done so and thus prevented OpenAI from extracting their data.
In parallel, this week in California, a class-action lawsuit alleging privacy violations by OpenAI was dismissed.