Rightsholders Can't Use OpenAI and Anthropic to Dismantle Meta's Seeding Defense
Meta previously argued that uploading pirated books to other BitTorrent users was an unavoidable side effect of downloading them, and therefore fair use. Rightsholders suing Meta in three related cases tried to dismantle this theory by asking OpenAI and Anthropic to reveal how they torrented shadow library data. A magistrate judge denied the request, but that's not the end of the torrenting dispute.
Over the past two years, rightsholders of all kinds have filed lawsuits against companies that develop AI models.
Meta is among a long list of companies now being sued for this allegedly infringing activity. This includes a class action lawsuit filed by authors including Richard Kadrey and Sarah Silverman, which accused Meta of training its Llama models on pirated books, and of sharing those books with other BitTorrent users in the process.
Last summer, Judge Vince Chhabria ruled that the AI training itself was fair use , leaving the BitTorrent distribution claims as the last live part of the case. Earlier this year, Meta added a new line of defense to those claims. In a supplemental interrogatory response, the company argued that any uploading of pirated books during its torrent downloads was part-and-parcel of a fair use purpose.
Meta stated that BitTorrent was a more efficient and reliable means of obtaining the datasets, and in the case of Anna s Archive the only way to get them in bulk.
Since torrent users upload to each other by design, any sharing was simply an inherent characteristic of the BitTorrent protocol.
Both of Meta s torrenting claims are now being tested in three related lawsuits , filed by Chicken Soup for the Soul, academic publisher Cognella, and John Carreyrou s Cambronne Inc. These three cases are all assigned to Judge Chhabria and target the same shadow library torrenting activity.
Instead of waiting for Meta to document the technical details of its torrent client setup, the publishers went to the two AI rivals that could potentially disprove the seeding requirement.
In August, they subpoenaed OpenAI and Anthropic for the identity, versions and configurations of every torrent client the companies have used since 2019. This specifically includes any records of efforts to prevent seeding.
In similar lawsuits, both companies have admitted that they used books from shadow libraries.