Facebook Apparently Trained Its AI by Torrenting Pirated Books Stolen From Authors – Casson Living – World News, Breaking News, International News

Recent court documents that have come to light suggest that Meta, formerly known as Facebook, may have engaged in the unethical practice of using pirated literature from Library Genesis for training its artificial intelligence systems. According to these documents, Mark Zuckerberg allegedly gave the green light for this controversial approach, raising significant legal and ethical questions.

Library Genesis, often referred to as LibGen, operates as a “shadow library,” providing free access to an extensive range of books, articles, and magazines. The revelation that a major corporation like Meta resorted to utilizing such pirated materials highlights the lengths to which some businesses may go to obtain data for their AI initiatives.

While Meta and other leaders in AI claim that such usage can fall under the umbrella of “fair use,” the ongoing litigation involving authors like Ta-Nehisi Coates and Sarah Silverman will ultimately clarify the validity of this assertion. The newly revealed documents also indicate that some Meta employees were aware of the questionable nature of their actions, with one engineer expressing concerns about downloading torrents on a corporate device.

The implication that Mark Zuckerberg was involved in endorsing the use of pirated content further complicates Meta’s position. Plaintiffs argue that this new information could severely weaken any defense the company might present, particularly in light of evidence suggesting that Meta actively participated in distributing the pirated books by seeding their torrents.

This situation brings to the forefront the ethical and legal challenges that the tech industry faces in its quest for data to enhance AI development. It remains uncertain how this controversy will shape the practices of Meta and other companies moving forward.