Overview
- U.S. District Judge Sidney H. Stein denied OpenAI’s bid to strike a distinct book‑download “shadow library” theory, finding prior complaints put the company on notice of reproduction allegations.
- The court held a reasonable jury could find ChatGPT outputs substantially similar to plaintiffs’ works, citing summaries and a proposed sequel outline tied to George R.R. Martin’s series.
- Stein struck references to newer systems such as GPT‑4V, GPT‑4.5, GPT‑5 and successors, narrowing the case to earlier models from GPT‑3 through GPT‑4o Mini, while rejecting Microsoft’s attempt to exclude GPT‑4o allegations.
- The litigation proceeds as an MDL in the Southern District of New York, and fair use remains unaddressed at this stage.
- With statutory damages up to $150,000 per work and discovery focusing on output logs, vendor book lists, and internal communications, parties point to the Anthropic case and its $1.5 billion settlement as contextual precedent.