Newspaper Publishers Demand Destruction of AI Models in Federal Scraping Lawsuit.

The Seattle Times and Newsday File Federal Copyright Lawsuit Against OpenAI and Microsoft

The Seattle Times and Newsday have jointly filed a federal copyright infringement lawsuit against OpenAI and Microsoft in the U.S. District Court for the Southern District of New York. The two regional news organizations allege that the tech giants systematically scraped paywalled journalism and copyrighted news archives without authorization to train commercial artificial intelligence models.

Demand for Data Destruction and Core Legal Claims

The legal filing aligns with high-profile intellectual property litigation previously initiated by major publishers, escalating the legal scrutiny surrounding AI training datasets:

• Injunction for Destruction of AI Models: The publishers are seeking a federal injunction requiring OpenAI and Microsoft to destroy all training datasets and LLM weights that incorporate their copyrighted news articles.

• Direct Content Reproduction: The lawsuit alleges that platforms such as ChatGPT, Microsoft Copilot, and Bing AI features reproduce verbatim excerpts and closely paraphrased summaries of proprietary investigative reporting.

• Threat to Subscription Revenue: Publishers argue that AI-generated search summaries and conversational answers reduce direct web traffic, undermining paid digital subscriptions and advertising revenue streams essential for funding local journalism.

• Executive Position: Alan Fisko, President and CEO of The Seattle Times, emphasized in an internal staff memo that the company invests millions of dollars annually to produce original journalism and must aggressively protect its intellectual property from unauthorized commercial exploitation.

Corporate Responses and Upcoming Court Schedule

In response to the filing, corporate representatives defended their training methodologies and partner engagement:

• OpenAI Defense: Representatives stated that their AI models are trained strictly on publicly available internet data under established Fair Use legal doctrines.

• Microsoft Statement: Microsoft representatives expressed surprise at the litigation, stating that the company values regional journalism and remains committed to finding collaborative resolutions to copyright disputes.

• Federal Court Timeline: The U.S. District Court for the Southern District of New York is scheduled to review the publishers' request for preliminary injunctive relief during a hearing set for September 25.

In copyright lawsuits involving generative AI, a request for a court-ordered destruction of trained models is known as "algorithmic destruction." Because large language models integrate training data into complex mathematical weights rather than storing raw text, removing specific copyrighted content requires retraining the model or wiping its entire set of checkpoints—a process that entails enormous financial costs.

The central legal battle hinges on whether the use of copyrighted material constitutes transformative "fair use" or commercial copyright infringement. While OpenAI and Microsoft argue that analyzing text patterns to learn language structures is legally protected, news organizations contend that generating summaries—which serve as substitutes for reading the original articles—undermines the economic value of their copyrighted works.

These lawsuits highlight a growing rift within the media industry: while some publishers choose to sue to protect their data archives, others are negotiating lucrative, long-term licensing agreements to supply real-time news content to AI models. The outcome of future proceedings will establish a critical legal precedent, shaping how AI companies compensate content creators moving forward.

 

 

💬 AI Content Assistant

Ask me anything about this article. No data is stored for your question.

The Seattle Times and Newsday File Federal Copyright Lawsuit Against OpenAI and Microsoft

The Seattle Times and Newsday have jointly filed a federal copyright infringement lawsuit against OpenAI and Microsoft in the U.S. District Court for the Southern District of New York. The two regional news organizations allege that the tech giants systematically scraped paywalled journalism and copyrighted news archives without authorization to train commercial artificial intelligence models.

Demand for Data Destruction and Core Legal Claims

The legal filing aligns with high-profile intellectual property litigation previously initiated by major publishers, escalating the legal scrutiny surrounding AI training datasets:

• Injunction for Destruction of AI Models: The publishers are seeking a federal injunction requiring OpenAI and Microsoft to destroy all training datasets and LLM weights that incorporate their copyrighted news articles.

• Direct Content Reproduction: The lawsuit alleges that platforms such as ChatGPT, Microsoft Copilot, and Bing AI features reproduce verbatim excerpts and closely paraphrased summaries of proprietary investigative reporting.

• Threat to Subscription Revenue: Publishers argue that AI-generated search summaries and conversational answers reduce direct web traffic, undermining paid digital subscriptions and advertising revenue streams essential for funding local journalism.

• Executive Position: Alan Fisko, President and CEO of The Seattle Times, emphasized in an internal staff memo that the company invests millions of dollars annually to produce original journalism and must aggressively protect its intellectual property from unauthorized commercial exploitation.

Corporate Responses and Upcoming Court Schedule

In response to the filing, corporate representatives defended their training methodologies and partner engagement:

• OpenAI Defense: Representatives stated that their AI models are trained strictly on publicly available internet data under established Fair Use legal doctrines.

• Microsoft Statement: Microsoft representatives expressed surprise at the litigation, stating that the company values regional journalism and remains committed to finding collaborative resolutions to copyright disputes.

• Federal Court Timeline: The U.S. District Court for the Southern District of New York is scheduled to review the publishers' request for preliminary injunctive relief during a hearing set for September 25.

In copyright lawsuits involving generative AI, a request for a court-ordered destruction of trained models is known as "algorithmic destruction." Because large language models integrate training data into complex mathematical weights rather than storing raw text, removing specific copyrighted content requires retraining the model or wiping its entire set of checkpoints—a process that entails enormous financial costs.

The central legal battle hinges on whether the use of copyrighted material constitutes transformative "fair use" or commercial copyright infringement. While OpenAI and Microsoft argue that analyzing text patterns to learn language structures is legally protected, news organizations contend that generating summaries—which serve as substitutes for reading the original articles—undermines the economic value of their copyrighted works.

These lawsuits highlight a growing rift within the media industry: while some publishers choose to sue to protect their data archives, others are negotiating lucrative, long-term licensing agreements to supply real-time news content to AI models. The outcome of future proceedings will establish a critical legal precedent, shaping how AI companies compensate content creators moving forward.