Meta Platforms is currently engaged in legal battles over allegations of copyright infringement related to the training of its artificial intelligence models. The company is contesting claims from both individual authors and Entrepreneur Media, asserting that its use of copyrighted materials falls under the doctrine of fair use.
In a lawsuit filed in July 2023, a group of authors, including Sarah Silverman, Richard Kadrey, and Christopher Golden, accused Meta of using their copyrighted works without authorization to train its large language model, Llama. The authors alleged that Meta sourced their works from unauthorized “shadow libraries” such as Library Genesis (LibGen) and Bibliotik, which host pirated books. They argued that Meta’s AI models could generate detailed summaries of their books, indicating that the AI had been trained on their copyrighted content. ([entrepreneur.com](https://www.entrepreneur.com/business-news/sarah-silverman-sues-openai-meta-for-use-of-copyrighted/455579?utm_source=openai))
Meta responded by invoking the fair use doctrine, contending that its AI training was transformative and did not replicate the authors’ books or serve as a substitute for reading them. The company emphasized that Llama was designed to assist with tasks like generating business reports, translating conversations, and composing letters, rather than reproducing the plaintiffs’ works. ([tradingview.com](https://www.tradingview.com/news/reuters.com%2C2025%3Anewsml_L6N3Q80KD%3A0-meta-says-copying-books-was-fair-use-in-authors-ai-lawsuit/?utm_source=openai))
In June 2025, U.S. District Judge Vince Chhabria dismissed the authors’ lawsuit, stating that they failed to present sufficient evidence that Meta’s AI would cause significant market harm to their works. The judge noted that the ruling did not establish the lawfulness of Meta’s use of copyrighted materials but rather highlighted deficiencies in the plaintiffs’ arguments. ([theguardian.com](https://www.theguardian.com/technology/2025/jun/26/meta-wins-ai-copyright-lawsuit-as-us-judge-rules-against-authors?utm_source=openai))
Separately, in November 2025, Entrepreneur Media filed a lawsuit against Meta, alleging that the company used pirated versions of its books and magazine articles to train its AI models. The complaint claimed that Meta downloaded and redistributed copyrighted works through torrenting networks, thereby becoming a distributor of pirated material. Entrepreneur Media argued that Meta’s actions constituted direct and contributory copyright infringement, as well as violations of the Digital Millennium Copyright Act. ([medianama.com](https://www.medianama.com/2025/11/223-entrepreneur-meta-pirated-books-llama-ai-models/?utm_source=openai))
Meta has consistently maintained that its use of publicly available datasets for AI training is legally permissible under the fair use doctrine. The company asserts that its AI models are designed to perform transformative tasks that do not infringe upon the market for the original works.
These legal disputes underscore the ongoing tension between AI development and copyright law, as courts continue to grapple with the implications of using copyrighted materials to train machine learning models. The outcomes of these cases may have significant ramifications for the future of AI training practices and the rights of content creators.