In a significant legal development, Meta, a company currently banned in Russia and classified as extremist, faced a lawsuit in 2023 regarding the training of its LLM model, Llama. The allegations center around the use of pirated content sourced from torrent trackers.
Until recently, Meta had been submitting edited documents to the court. However, a judge has now mandated the release of original documents, shedding light on the internal discussions within the company. In one conversation, an engineer expressed concerns about the appropriateness of using torrents on a corporate laptop, indicating that the company did indeed utilize pirated content for its artificial intelligence training. Another exchange hints at the possibility that Mark Zuckerberg, referred to as “MZ,” may have approved the use of such materials.
Among the sources of pirated content was LibGen, a vast Russian repository of pirated books, journals, and academic articles established in 2008. Meta reportedly also drew from various other libraries.
The company defends its actions by invoking the legal doctrine of “fair use,” which permits the application of copyrighted content without permission under specific circumstances. This defense will likely be a focal point in the ongoing legal proceedings.
