Court Documents Reveal OpenAI Knew Using Pirated Books Was Illegal and Harmful to Authors

By the editors ／ Model: gemma-4-26b-a4b-it-qat-6bitAs of: 2026-09-28

New court documents released in the lawsuit filed by the Authors Guild against OpenAI and Microsoft reveal that executives at both companies were aware that using pirated books for AI training was illegal and posed a threat to the livelihoods of human authors.

The filings, submitted by the Authors Guild, detail that OpenAI and Microsoft leadership had knowledge of the use of "mass book piracy" to train their models. According to the documents, OpenAI's leadership was aware of the use of LibGen, a large-scale piracy library, as early as 2019. Internal communications suggest that while some employees raised concerns about the legal and public perception of using such datasets, the focus remained on avoiding negative publicity rather than ensuring legality.

The documents also include testimony regarding the potential impact of AI on the writing profession. Former OpenAI policy director Jack Clark reportedly stated in 2020 that AI developments could lead to systems that replace human labor, specifically mentioning concerns from genre novelists. Additionally, a researcher at OpenAI, Tarun Gogineni, is quoted in the filings as discussing the ability of GPT models to "automatically complete" works by authors such as George R.R. Martin, even after the authors' passing.

In response to the findings, Mary Rasenberger, CEO of the Authors Guild, stated that the filings show a "shocking disdain for writers," alleging that the companies chose to intentionally steal books rather than pay for them, despite knowing the potential for economic disruption and the degradation of culture. The legal proceedings, part of a class action case, are expected to move toward a hearing in early 2027.

Sources

Related