NYT v. OpenAI Copyright Verdict Expected

What this countdown tracks

Expected verdict in The New York Times Company v. Microsoft Corporation and OpenAI, the first major US copyright lawsuit over generative-AI training data to reach a jury. Filed on 27 December 2023 in the US District Court for the Southern District of New York and consolidated with parallel actions from the Daily News, the Center for Investigative Reporting and the New York Daily News, the case is on track for a verdict around early December 2026 after a fall 2026 trial date set by Judge Sidney Stein.

About the case

The Times alleges that OpenAI and Microsoft copied millions of New York Times articles, archives and Wirecutter reviews — without licence and at industrial scale — to train GPT-3.5, GPT-4, GPT-4-Turbo, GPT-4o and successor models, and that ChatGPT and Bing Chat output near-verbatim regurgitations of paywalled Times content on demand. The complaint pleads:

• Direct copyright infringement for unauthorised reproduction during training.
• Vicarious and contributory infringement against Microsoft for hosting and distributing models built on the corpus.
• DMCA Section 1202 claims for stripping copyright-management information.
• Common-law unfair competition.

OpenAI and Microsoft assert fair use under the four-factor test, arguing training is a transformative statistical process — not a substitutional copy — and that ChatGPT's output is, in normal use, paraphrased rather than verbatim.

Stakes

A plaintiff verdict could expose OpenAI and Microsoft to statutory damages of up to $150,000 per wilfully infringed work — a number that, multiplied across the Times' archive, runs into the tens of billions. It would also rewrite the licensing economics for every frontier-model lab in the world. A defence verdict — or a finding of fair use — would crystallise training-on-the-open-web as a legally defensible practice and remove a structural overhang from AI valuations.

Past results

The judge has already denied OpenAI's motion to dismiss most claims and ordered preservation of training-data logs that OpenAI had been deleting on a rolling basis. Discovery has produced internal documents on data-scraping pipelines, model-evaluation benchmarks against Times content, and licensing negotiations OpenAI conducted with other publishers (Axel Springer, AP, FT, News Corp).

How to follow

PACER docket 1:23-cv-11195. Reuters, Law360, Politico, the Times itself and the Verge cover hearings; recorded transcripts are filed within days. The New York Times has been reporting against its own legal interest, which has produced unusually candid contemporaneous coverage.

Related countdowns

Pair with Musk v. Altman trial, the SAG-AFTRA contract expiry, the EU AI Act GPAI rules going live and the GPT-6 launch window.

FAQ

When is the verdict expected? Around early December 2026, following a fall 2026 trial. Who is the judge? Judge Sidney H. Stein (SDNY). What is the maximum exposure? Statutory damages up to $150,000 per wilfully infringed work. What is the core defence? Fair use under 17 U.S.C. § 107 — that training is transformative and non-substitutional.