OpenAI is facing growing scrutiny and potential regulatory penalties after deleting datasets containing copyrighted books that were reportedly used to train its AI models. The move has raised concerns among rights holders and legal experts, who argue that the company’s actions may be an attempt to avoid accountability rather than a genuine effort to comply with copyright law. The datasets in question included large volumes of books that were scraped from the internet without authorization from publishers or authors. These materials were allegedly used to train models like GPT-3 and GPT-4, forming the foundation of OpenAI’s language capabilities. After reports surfaced about the existence of these datasets, OpenAI removed them from public access, but did not provide a detailed explanation for the deletion or clarify whether the data was ever properly licensed. Legal experts warn that the sudden removal of the datasets—without transparency—could be interpreted as an effort to obstruct investigations or evade liability. In several jurisdictions, including the United States and the European Union, companies can face significant fines if they fail to demonstrate compliance with copyright laws, especially when using copyrighted material for commercial AI training. The deletion has intensified pressure on OpenAI to clarify its data sourcing practices. Critics argue that the lack of transparency undermines trust and could lead to more aggressive regulatory action. Publishers and authors have long voiced concerns about the unauthorized use of their works in AI training, and this incident has reignited debates over intellectual property rights in the age of generative AI. OpenAI has not publicly addressed the specifics of the deletion or confirmed whether the data was used in any of its current models. The company has previously stated that it uses a mix of licensed and public data, but has not provided a full accounting of its training data sources. As legal challenges and government inquiries into AI training practices grow, OpenAI’s handling of this situation may have lasting implications for how AI companies manage data rights and compliance.