Meta accused of using ‘thousands of pirated books’ in artificial intelligence training

Meta got another one Case regarding unauthorized use of books in the training of the Lama language modelpowering the latest productive AI tools. The action, launched on Wednesday (2) in the USA, was suggested by author Christopher Farnsworth.

According to the novelist, the technology giant Used “thousands of pirated books” to train chatbots respond to user requestsAs reported by Reuters. The material is part of a database of around 800GB allegedly used by owner Facebook.

The author claims that this open access database titled “Books3” contains works obtained illegally. He says he doesn’t know the exact origin of the books, but says the company led by Mark Zuckerberg I didn’t pay to use these in AI training.

This content will include books by Farnsworth, who wrote titles such as: Flashmob, Killfile And President’s Vampireamong others. The author requested compensation from the court for an undisclosed amount. and wants Meta to be forced to stop using the material immediately.

transaction stacks

In addition to this new process, where he is accused of using thousands of pirated books in artificial intelligence training, Meta has been the target of many other similar actions Lately. One of them has been run by actress, comedian and writer Sarah Silverman since the middle of last year.

According to Silverman, the controller of Instagram and WhatsApp is Large amounts of copyrighted material to feed AI bots. The author claims that leaked documents about the Big Tech technology training process show that the company did not seek permission to use the content.

In the lawsuit, which also includes OpenAI as a defendant, the author demands compensation for the misuse of his works. Already companies defend themselves by claiming “fair” use of material and claims that the lawsuits threaten advances in the AI ​​industry.