OpenAI Copyright Lawsuit: Newspapers Demand Model Deletion

The Seattle Times and Newsday filed suit against OpenAI and Microsoft on 6 September 2026, demanding the companies delete AI models trained on their reporting. If courts order model destruction, it could force every AI tool, including wallet assistants, to rethink where their answers come from.

What actually happened

According to The Verge, the two publishers claim OpenAI trained its models on their articles without consent, and that those models sometimes repeat article text word for word when users ask questions. Microsoft is named as a defendant because its Copilot chatbot runs on OpenAI's technology. The lawsuit asks courts to order the destruction of any stored copies of the papers' work, plus the training data and the models built from it. This filing joins nearly 400 local newspapers already suing the same two companies over lost traffic and subscription revenue. OpenAI and Microsoft had not responded to requests for comment as of the report.

How we got here

This is not an isolated case. The New York Times, Ziff Davis, Merriam-Webster, and Encyclopedia Britannica have filed similar copyright claims against OpenAI, per The Verge. The nearly 400 local newspapers already suing argue that chatbot answers replace visits to their sites, cutting ad and subscription income. The new complaint lands just two days after OpenAI began rolling out GPT-6 Astra, its most capable model yet, on 4 September 2026. That timing puts new pressure on how fresh, more powerful models source their training material.

Why this matters for you

For everyday wallet app users, this case could shape how trustworthy AI answers feel inside the tools you already use. If courts force retraining or licensing deals, chatbots may quote less news directly and link back to original sources instead. For builders integrating AI assistants into wallets or bonuz-style super apps, this raises the cost and legal risk of using unlicensed content to train features. It also signals that provenance of data, not just model performance, will matter more as AI gets embedded into daily financial tools.

The bigger question

No court has ever ordered a major AI model deleted for copyright reasons. If one does here, could that ruling actually be enforced across cloud servers, enterprise apps, and consumer devices already running that model? Or would destruction remain mostly symbolic once a model is this deeply embedded in everyday products?

What to watch

Watch for OpenAI and Microsoft's formal legal response, not yet filed as of 6 September 2026. Track whether more publishers join the nearly 400 newspapers already suing, and whether courts merge these cases with The New York Times lawsuit. Any ruling on deleting a trained model would set a benchmark for every AI product built afterward, wallets included.