When was the NYT v. OpenAI lawsuit filed?

The complaint was filed on December 27, 2023 in the Southern District of New York, case number 1:23-cv-11195.

How much is the NYT asking for?

The complaint seeks 'billions of dollars of statutory and actual damages' plus destruction of GPT models trained on Times works. Exact damages will be determined at trial.

No trial date is set. Summary judgment briefing is scheduled to conclude April 2, 2026. If any claims survive summary judgment, trial is likely late 2026 or 2027.

Has there been any ruling yet?

Yes — Judge Stein denied OpenAI's motion to dismiss in March 2025, allowing direct and contributory infringement claims to proceed. Judge Stein also affirmed Magistrate Judge Wang's order compelling OpenAI to produce 20 million ChatGPT logs in January 2026.

New York Times v. OpenAI & Microsoft: Case Tracker

Q: What is OpenAI's defense?

OpenAI argues its use of Times articles is fair use, asserts that memorization is a 'rare bug' not a feature, and has raised statute-of-limitations defenses for certain training-era claims. Judge Stein rejected several of these arguments on the motion to dismiss.

Case summary

Plaintiff: The New York Times Company
Counsel: Susman Godfrey LLP (Steven Lieberman, Ian Crosby)
Defendants: OpenAI, Inc.; OpenAI GP, LLC; OpenAI OpCo, LLC; OpenAI Global, LLC; Microsoft Corporation
Counsel: Latham & Watkins; Morrison & Foerster
Court: U.S. District Court, Southern District of New York
Case number: 1:23-cv-11195
Filed: December 27, 2023
Judge: Hon. Sidney H. Stein (District Judge)
Hon. Ona T. Wang (Magistrate)
Claim types: Direct & contributory copyright infringement; DMCA §1202; common-law unfair competition (dismissed); trademark dilution
Damages sought: "Billions of dollars" — statutory and actual
Current status: Active discovery; SJ briefing concludes April 2, 2026
Last verified: Apr 24, 2026 · Editor: Tom A.

TL;DR

In December 2023, The New York Times sued OpenAI and Microsoft alleging they copied millions of Times articles — without license or payment — to train the GPT family of large language models that power ChatGPT and Bing Copilot. The Times's complaint included exhibits showing ChatGPT reproducing near-verbatim passages from Times articles in response to user prompts, along with fabricated quotes attributed to the paper.

OpenAI moved to dismiss, arguing fair use, that memorization is a "rare bug," and that training-era claims from 2019–2020 are time-barred. In March 2025, Judge Sidney Stein denied most of OpenAI's motion, letting direct and contributory infringement claims proceed while trimming some DMCA and unfair-competition counts. The case is now in contentious discovery: in January 2026, Judge Stein affirmed Mag. Judge Wang's order compelling OpenAI to produce 20 million de-identified ChatGPT logs — a landmark discovery ruling. Summary judgment briefing concludes April 2, 2026.

Why it matters

Sets the U.S. precedent for news-publisher AI claims. If the Times wins on liability, every major newspaper has a template to file; if OpenAI wins on fair use at scale, the copyright theory collapses and licensing leverage disappears.
The 20 million-log discovery order rewrites AI discovery norms. Courts can now compel production of user-chat logs to probe training and output behavior — a fact pattern every generative-AI defendant will face.
Damages exposure is existential. Statutory damages for willful infringement run up to $150,000 per work. Millions of articles multiplied by that ceiling is not a rounding error, even for OpenAI and Microsoft.
Judge Stein's rulings diverge from early California fair-use rulings in Kadrey v. Meta and Bartz v. Anthropic on the fair-use balance — setting up a clean circuit split posture if it reaches appeal.

Timeline of events

Dec 27, 2023

Complaint filed

The Times files 69-page complaint in SDNY alleging direct and contributory copyright infringement, DMCA violations, unfair competition, and trademark dilution against OpenAI and Microsoft.
Feb 26, 2024

OpenAI motion to dismiss

OpenAI moves to dismiss certain claims, arguing fair use, that memorization is a rare bug, and that 2019–2020 training claims are outside the three-year statute of limitations.
Apr 30, 2024

NYT amends complaint

Amended complaint adds additional exhibits of GPT-4 reproducing Times content, including fabricated quotes attributed to Times reporters.
Mar 26, 2025

Order on motion to dismiss

Judge Stein partially grants and partially denies OpenAI's motion — direct and contributory infringement survive; certain DMCA and common-law unfair competition claims dismissed.
Apr 04, 2025

Judge Stein issues written opinion

43-page opinion explains the MTD ruling — emphasizes "numerous and well-publicized" instances of ChatGPT generating Times-derived content as grounds for contributory liability.
Nov 07, 2025

MJ Wang orders 20M-log production

Magistrate Judge Wang compels OpenAI to produce the entire de-identified 20 million ChatGPT log sample — including logs not implicating Times works — as relevant to OpenAI's fair-use defense.
Dec 05, 2025

Reconsideration denied

MJ Wang denies OpenAI's motion for reconsideration of the 20M-log order.
Jan 05, 2026

Judge Stein affirms discovery order

District court denies OpenAI's Rule 72 objections and affirms MJ Wang's 20M-log discovery order — "not clearly erroneous or contrary to law."
Apr 08, 2026

OpenAI 30(b)(6) witness deemed unprepared

Judge gives news organizations and authors additional deposition time after finding OpenAI's corporate designee lacked preparation on noticed topics.
Apr 02, 2026

SJ briefing scheduled to conclude

Expert reports were due November 14, 2025; summary judgment briefing closes April 2, 2026 per the scheduling order.

Key documents

PDFs hosted via CourtListener / RECAP where available. Dkt. numbers correspond to the SDNY docket.

Complaint

Dkt. 1 — 69 pages

12/27/23

Amended complaint

Dkt. 40

04/30/24

OpenAI motion to dismiss

Dkt. 52

02/26/24

Order on motion to dismiss

Dkt. 342

03/26/25

Stein opinion (MTD explanation)

Dkt. 362 — 43 pages

04/04/25

Stein order affirming 20M-log production

Dkt. 972

01/05/26

Rulings & outcomes

Motion to dismiss denied in part

Mar 26, 2025 · Stein, J.

Holding: Direct and contributory copyright infringement claims survive. DMCA §1202 claims dismissed with leave to amend. Common-law unfair competition claim dismissed. Statute-of-limitations defense for 2019–2020 training rejected.

The Times has plausibly alleged numerous and well-publicized instances of ChatGPT generating content derived from its copyrighted articles that warrant continuation of its direct and contributory infringement claims.
— Op. at 14

OpenAI ordered to produce 20M ChatGPT logs

Nov 07, 2025 · Wang, M.J. (aff'd Jan 05, 2026)

Holding: OpenAI must produce the entire de-identified 20 million ChatGPT log sample — including logs unrelated to Times works — because those logs are relevant to OpenAI's fair-use defense. District court affirmed on Rule 72 review.

Output logs that do not contain reproductions of News Plaintiffs' works may still be relevant to OpenAI's fair use defense.
— Dkt. 896 at 3

Additional deposition time granted after 30(b)(6) failure

Apr 08, 2026 · Wang, M.J.

Holding: News organizations and consolidated author plaintiffs given additional time for further depositions after court found OpenAI's corporate designee was unprepared on noticed topics. OpenAI ordered to re-designate and produce a properly prepared witness.

Money at stake

Claimed damages

"Billions"

Statutory + actual

Statutory ceiling per work

$150,000

17 U.S.C. § 504(c) — willful

Awarded to date

Pre-judgment

The Times also seeks a permanent injunction and "destruction" of GPT models trained on Times works under 17 U.S.C. § 503(b) — a remedy that, if granted, would have industry-wide implications for deployed LLMs.

Expert commentary

Stein's ruling allows The Times to continue pursuing claims that OpenAI's outputs included copyrighted content and resulted in user infringement, which diverges from the stance of California judges who have dismissed similar allegations.
— Reuters Legal, Apr 4, 2025

Last week saw another flood of discovery letter motions submitted, mostly by Defendants. OpenAI moved to compel The New York Times to produce any outputs which The Times alleges infringe its asserted works or cause market harm.
— McKool Smith AI Litigation Update, Oct 20, 2025

Press coverage

Reuters LegalJudge explains order for New York Times in OpenAI copyright case
Ars TechnicaNYT v. OpenAI order (Jan 5, 2026) — full PDF
Law360Case coverage hub — The New York Times Company v. Microsoft Corporation et al.
McKool SmithAI Infringement Case Updates, Oct 20, 2025
The New York TimesThe Times sues OpenAI and Microsoft over A.I. use of copyrighted work
Bloomberg LawNYT-OpenAI case tracker (subscription)

Frequently asked questions

What is the NYT v. OpenAI lawsuit about?

The New York Times sued OpenAI and Microsoft in December 2023, alleging they used millions of Times articles without permission to train GPT models that power ChatGPT and Bing Copilot. The Times claims direct and contributory copyright infringement, DMCA violations, and trademark dilution, and seeks "billions of dollars" in statutory and actual damages plus destruction of infringing models.

What is OpenAI's defense?

OpenAI argues its use of Times articles is transformative fair use under 17 U.S.C. § 107, that memorization of training data is a "rare bug" rather than a product feature, and that certain training-era claims from 2019–2020 fall outside the three-year statute of limitations. Judge Stein rejected most of these arguments on the motion to dismiss in March 2025.

How much is The Times asking for?

The complaint seeks "billions of dollars of statutory and actual damages." Under 17 U.S.C. § 504(c), statutory damages for willful infringement run up to $150,000 per work. With millions of Times articles at issue, the theoretical ceiling is enormous — though final damages (if any) will be determined at trial.

When is the trial?

No trial date is set. Expert reports were due November 14, 2025. Summary judgment briefing is scheduled to conclude April 2, 2026. If claims survive summary judgment, trial is likely late 2026 or 2027.

What was the 20 million log order?

In November 2025, Magistrate Judge Wang ordered OpenAI to produce a de-identified sample of 20 million ChatGPT conversation logs — including logs that do not reproduce Times works — holding they are relevant to OpenAI's fair-use defense. Judge Stein affirmed that order on January 5, 2026. It is one of the most significant discovery rulings in any AI case to date.

How is this different from Bartz v. Anthropic?

Bartz v. Anthropic (NDCA) involved authors suing Anthropic over pirated books used for Claude training. Anthropic settled for $1.5 billion. NYT v. OpenAI involves a different plaintiff class (news publishers, not authors) in a different circuit (SDNY), with rulings that have been more favorable to plaintiffs than NDCA fair-use decisions. The cases are on parallel tracks and could produce a circuit split.