Seattle Times and Newsday Sue OpenAI and Microsoft Over AI Training

The copyright conflict between generative AI companies and news publishers is expanding. The Seattle Times and Newsday have sued OpenAI and Microsoft, alleging that OpenAI used their journalism to train AI models without permission and that the systems sometimes reproduce passages from their reporting when answering users.

What the publishers allege

The complaint, as described in the source material, centers on several claims:

• OpenAI used the publishers’ articles as training material without obtaining a license;
• its models may output passages that are identical or closely similar to published reporting;
• Microsoft should also be liable because Copilot is built on OpenAI technology;
• chatbots can reduce the need for users to visit news websites for reporting and answers, potentially affecting subscriptions and other revenue.

The requested remedies go beyond financial compensation. The two publishers are seeking the destruction of copies of their works held by the companies, along with training datasets and AI models that incorporate those works. If such a request becomes a major focus of the case, the dispute would concern not only whether the original use was infringing, but also what should happen to models after training has already taken place.

Part of a broader legal fight

The lawsuit is not an isolated action against OpenAI. The source notes similar cases brought by The New York Times, Ziff Davis, Merriam-Webster, and Encyclopaedia Britannica. It also points to a recent lawsuit involving nearly 400 local newspapers against OpenAI and Microsoft. Those publishers argue that chatbots can provide reporting and answers directly, reducing visits to their sites and putting valuable subscription revenue at risk.

OpenAI and Microsoft had not immediately responded to a request for comment at the time of publication. The available account therefore primarily reflects the allegations described in the lawsuit and the publisher-side concerns. The court will still need to determine whether the claims are valid and how responsibility should be divided.

Why the case matters

The dispute highlights an unresolved question in the AI economy: how should content creators be compensated when their work helps build and operate generative systems? For news organizations, articles are not only copyright assets. They also attract readers to websites, newsletters, accounts, and subscription products. If an AI assistant delivers a useful answer without sending the user to the original publisher, the outlet may lose both attention and a commercial conversion opportunity.

For AI companies, the litigation could affect how training data is acquired, how licensing agreements are negotiated, and how model outputs are monitored. Future arguments may focus on more than whether public material can be used during training. They may also examine whether a system reveals recognizable source text, whether answers should provide attribution and links, and what duties companies have to manage data and output risks.

The case may therefore become another important test of the relationship between journalism copyright, AI training practices, and the business model that supports digital news.

Source: The Verge AI