#

Key takeaways

• New unredacted filings added detail to the copyright dispute.

• The report centers on internal statements about training data, paywalls, and publisher traffic.

• The underlying exhibits remain sealed.

• The legal question about using copyrighted material for model training is still unresolved.

• The source reports no Morocco-specific facts.

What the report says

TechCrunch reported on September 17, 2026 that newly unredacted filings added more detail to The New York Times' copyright lawsuit against OpenAI and Microsoft. The report says the new material includes internal statements about AI training data, paywalls, and the effect of AI answer products on publisher traffic.

The article also says much of the new information came from the Times' own brief. It notes that the underlying exhibits remain sealed. That means the public record is still incomplete.

Why the filings matter

The filing details appear to sharpen the dispute around how AI systems use published material. The report focuses on the tension between training data practices and copyright claims. It also points to concerns about how AI answer products may affect traffic to publishers.

At the same time, the report does not resolve the central legal issue. The question of whether copyrighted material can be used to train models remains open. The article presents this as an unresolved legal matter, not a settled outcome.

What remains unknown

The source is careful about limits. It says the exhibits are still sealed, so readers do not have the full evidentiary record. It also does not provide a final ruling or a legal conclusion.

That matters because the new filings add context, but not closure. The dispute still depends on what the court accepts, what remains under seal, and how the legal arguments develop.

How to read this development

This is best understood as a procedural update with substantive detail. The filings may help explain the parties' positions, but they do not end the case. They also do not settle the broader debate over AI training and copyright.

For readers following AI policy and publishing disputes, the main point is simple. More detail has entered the record, but the core question is still open.

Morocco relevance

The source reports no Morocco-specific facts. As a conditional global lesson, readers should watch how unresolved copyright disputes can shape AI product design and publisher relationships.

Bottom line

The unredacted filings add texture to an already important case. They highlight training data, paywalls, and traffic concerns. But the legal question at the center of the dispute remains unresolved.