The New York Times filed a motion to amend its copyright complaint against OpenAI and Microsoft, alleging that Microsoft actively encouraged OpenAI to infringe on its copyrights. The amended complaint claims Microsoft built a bespoke supercomputing system specifically for training artificial intelligence on Times content without permission.

Graham James, a spokesperson for The New York Times, stated the organization asked the court for permission to file the amended complaint to strengthen its case. "Today, we asked the court for permission to file an amended complaint that further strengthens our case, clarifying our claim of contributory infringement against Microsoft based on new law and new evidence uncovered during discovery," James said. "As we have long alleged, Microsoft actively encouraged OpenAI to steal our copyrighted works."

The amended complaint argues that the supercomputer was constructed with the explicit purpose of training AI on copyrighted works without authorization. It also alleges that Microsoft designed the system to utilize internet content curated to feature Times works disproportionately for large language model training. The filing claims Microsoft's deployment of Times-trained large language models contributed to a one trillion dollar increase in its market capitalization over the past year.

A Microsoft spokesperson described the amended complaint as "a last-ditch effort by the plaintiff to save its claim from unfavorable precedent set in other recent rulings," while OpenAI spokesperson Drew Pusateri stated, "Our models empower innovation, are trained on publicly available data, and are grounded in fair use."

The newspaper sued OpenAI in 2023, alleging that ChatGPT was trained on its articles without permission and infringed on copyrights by outputting articles verbatim. The initial complaint claimed ChatGPT caused market harms by positioning itself as a substitute for a subscription and that outputs summarizing Wirecutter reviews resulted in lost affiliate link commissions. The filing cited examples of users bypassing paywalls by requesting additional paragraphs of articles from ChatGPT and alleged that models outputted several paragraphs of content without specific user prompts.

The paper asserted that Bing Chat cited fake quotes from Moira Forbes and that ChatGPT fabricated an article linking non-Hodgkin's lymphoma to consuming orange juice, attributing it to The New York Times. OpenAI has argued that ChatGPT does not substitute for a Times subscription because it transforms the material for a different use. A federal judge stated in June that proving market harms could be a successful argument against AI training on copyrighted works.