Background
U.S. News & World Report, L.P. sues multiple OpenAI entities, alleging that U.S. News content was copied and used to build training datasets and train GPT models, and that OpenAI products then output uncredited and uncompensated material that can substitute for U.S. News offerings and harm its reputation, including by attributing incorrect information to U.S. News. The complaint also emphasises U.S. News rankings as a distinctive content category, describing them as the product of proprietary methodologies, data collection, and publication protocols tied to U.S. News brand trust. The pleading notes that many earlier lawsuits are currently consolidated in the same court, situating this action within the wider SDNY OpenAI copyright litigation environment without, on its face, proving the procedural steps that later occurred on this specific member docket [#1].
AI Interaction
The dispute is framed around large language models and the end to end pipeline, namely collection or scraping of online content into training datasets, model training and storage of copied material, and downstream generation that can reproduce or closely track protected text in user-facing outputs [#1]. The complaint also pleads misattribution risk, alleging that OpenAI outputs can present inaccurate information while crediting U.S. News, creating an AI-mediated reputational harm pathway distinct from pure copying allegations [#1].
Notes:
The 16 January 2026 Order [#51] records rulings made at the 15 January 2026 discovery status conference, including directions to meet and confer and to report in a joint status letter, and it orders OpenAI to amend PII redactions and re produce a sample of output log data by 22 January 2026.