Plaintiffs defend expert testimony in OpenAI copyright MDL, citing Copilot display and notice-removal figures

What happened

On 9 November 2026, the plaintiffs in *In re OpenAI, Inc. Copyright Infringement Litigation* (No. 1:25-md-03143, S.D.N.Y.) filed their amended opposition to the defendants' Rule 702 motion to exclude expert testimony (Doc. 2501). The filing defends the plaintiffs' technical and industry experts against OpenAI's and Microsoft's bid to keep their testimony from the jury.

The two figures at the centre

The opposition cites expert Goldstein's analysis of Microsoft's Copilot logs: Copilot "systematically displays grounding content in outputs to the user in approximately 20% of conversations", which the plaintiffs say contradicts Microsoft's argument that web content used for grounding "is not revealed to the user". Goldstein also concludes Microsoft "likely copied substantial portions" of the news plaintiffs' articles as part of the retrieval-augmented generation process.

The opposition also cites expert Delorey's conclusion that OpenAI stripped Ziff Davis copyright notices from its training records "in 99.81% of the instances" — a conclusion the plaintiffs say "did not change and is not challenged" by the defendants.

What this does not decide

This is a filing by the plaintiffs, not a ruling. The court has not decided whether the experts may testify; the figures are the plaintiffs' experts' claims, offered to defend admissibility. Reply briefing in the MDL remains due 20 November 2026.