Apple sued by neuroscientists over using copyrighted books to train its AI model

Apple is facing a lawsuit from a pair of neuroscientists who claim the tech giant had used thousands of copyrighted books to train its AI model.

The lawsuit was filed in in the US District Court for the Northern District of California on Thursday 9th October by Susana Martinez-Conde and Stephen Macknik, neuroscientists at SUNY Downstate Medical Center in New York, Reuters reports.

They claim that Apple used “shadow libraries” to illegally access thousands of pirated books to train Apple Intelligence.

“Apple used illegally copied books, including materials from an online piracy repository called the ‘Shadow Library,’ as datasets to train its AI,” the neuroscientists said, adding that their books, entitled Champions of Illusion: The Science Behind Mind-Boggling Images and Mystifying Brain Puzzles and Sleights of Mind: What the Neuroscience of Magic Reveals About Our Everydayals, were among those used.

“Apple’s market value surged by over 200 billion dollars on the day it announced ‘Apple Intelligence,’ an AI-integrated system trained through such illegal means.”

Martinez-Conde and Macknik asked the court to order Google to pay unspecified damages and stop using copyright-infringing materials in AI training.

The case is just another in a long string of legal actions taken against tech giants who are just a little bit too eager to advance their AI models. Just in September, Apple was sued by authors Grady Hendrix and Jennifer Roberson for using its Applebot to reach shadow libraries and train its AI model on illegally shared books without the authors’ consent.

Similarly, Meta was sued for amassing at least 81.7 terabytes of data across multiple shadow libraries to train its Llama AI. Yet, the US district judge Vince Chhabria ruled that the use of those works by Meta is considered “fair use”.

And other companies are taking more legal – just slightly unusual methods to advance their AI. In June, it became known that artificial intelligence company, Anthropic, had cut millions of books from their bindings, scanned them to turn them into digital files, and discarded them after use, essentially destroying millions of books in the process.

In that case, the judge also ruled that the use of those books by Anthropic fell under the “fair use” category, because the company first legally bought the books and later destroyed them, leaving only digital files that weren’t distributed publicly.

AI companies need high-quality material to train their models, although obtaining a license and writer’s consent is often a long and costly procedure, pushing firms to finding workarounds.