“The scale is amazing”: Meta illegally distributed files on torrents on torrents, said copyright holders

Writers and copyright owners filed a lawsuit against Meta✴. They argue that recently opened internal emails of the company’s employees provide significant evidence of a large -scale violation of META✴ copyright. According to the plaintiffs, the company taught its own models of AI on pirate books and participated in a giant illegal torrent company, which includes tens of millions of works.

Image source: unsplash.com

In the statement of claim, copyright owners argue that Meta✴ participated in the torrent distribution “at least 81.7 terabytes from several shadow libraries through the Anna’s Archive website, including at least 35.7 data from Z-Library and Libgen, [… ] The scale of the illegal circuit of the META✴ torrents is amazing. ” In the statement of claim, it emphasizes that “much smaller acts of data piracy are only 0.008 percent of the number of copyrights protected by the copyright that Meta✴ pirates have copied” led to criminal prosecution.

Meta✴ for a long time resisted attempts to disclose information about the violation by the company copyright and participation in illegal torrents. In a petition for the deviation of the claim, the company claimed that the training of AI at the Libgen libraries is “conscientious use”. The company’s lawyers told the court that “the plaintiffs do not refer to any case in which any part of any book was actually loaded with the third party with Meta✴ through the torrent, not to mention the fact that the books of the plaintiffs were spreading in any way Meta✴ “.

Nevertheless, internal emails of the copyright owners were the internal letters of the META✴-research engineer Nikolai Bashlykov, in which he expresses concern about the use of corporate computers and IP addresses of META✴ “to load pirate content through torrents.” In September 2023, after consulting with lawyers, Bashlykov again tried to draw attention to the problem, emphasizing in the letter that “the use of torrents would entail the” distribution “of files, that is, the distribution of content outside the company, which may be legally unacceptable.”

The warnings of Bashlykov, apparently, were not heard, since Meta✴ continued to load and distribute terabytes from several shadow libraries until April 2024, taking some steps to mask their activities. Presumably, Meta✴ tried to hide the distribution without using Facebook servers when loading data set in order to avoid tracking the distribution source. The company also allegedly changed the settings of a torrent client to minimize the amount of information distributed.

Since the limited disclosure of information is currently ongoing, the company no longer disputes the fact of a violation of copyright. For META✴, the situation is now complicated by accusations of participating in the mass spread of pirate content. Nevertheless, the representative of Meta✴ stated in court that the company plans to “establish facts and debunk this baseless accusation in the order of simplified proceedings.”