NYT v. OpenAI (Copyright MDL): Billions Mass Tort

The New York Times, the Daily News and other publishers accuse OpenAI and Microsoft of copying millions of articles to build ChatGPT, in the lead case of the consolidated OpenAI copyright MDL. With statutory damages potentially in the billions, the case has erupted into a sanctions fight over whether OpenAI concealed its ability to search its own training data.

LATEST DEVELOPMENT · JUL 9, 2026

The Times, the Daily News and 15 other media plaintiffs moved for sanctions, accusing OpenAI of falsely denying it could search its training data and chat logs for their works, citing an internal ‘Project Giraffe’ detection tool and a 78-million-conversation database, and of deleting billions of ChatGPT outputs. They ask the court to exclude the 20-million-log sample, deem mass regurgitation established, and award fees.

PREDICTED OUTCOME

AI-GENERATED · METHODOLOGY →

Settlement ~65% likely HORIZON: settlement most likely 2027 to early 2028 CONFIDENCE: MEDIUM

The most likely ending is a negotiated resolution combining cash and a forward licensing deal, roughly 65 percent, signed between mid-2027 and early 2028 after bellwether rulings but before a jury verdict on the merits. Judge Stein’s March 2025 order preserving direct and contributory infringement, plus the affirmed 20-million-log production and now a credible spoliation and candor record, gives the publishers exactly the leverage that turns AI copyright cases into licensing transactions, and Anthropic’s $1.5 billion author settlement set the benchmark. Expect the sanctions motion to produce a middle outcome, fees, a narrowed adverse inference or an evidentiary presumption on regurgitation, rather than case-ending relief, which is the usual path for discovery-misconduct fights and which further prices the deal upward. A global or near-global resolution across the news-publisher track in the $500 million to $3 billion range, with per-outlet content licenses and attribution terms, is the realistic shape, with the Times plausibly taking a nine-figure cash component plus a multi-year license. OpenAI needs finality ahead of capital raising and cannot risk statutory damages across millions of registered articles, while Microsoft has both the cash and the precedent aversion to buy out its exposure separately and earlier. The leading alternative, around 20 percent, is that Stein rules training on lawfully acquired text substantially fair use and confines liability to output regurgitation, producing a materially split judgment and a much cheaper resolution. A clean defense win that ends the case, or a plaintiff verdict tried to judgment and affirmed, each remain well under 10 percent.

Core infringement claims survived dismissal before Judge Stein · Sanctions and spoliation record boosts publisher leverage · Anthropic $1.5B author settlement sets benchmark · OpenAI needs finality for fundraising and existing licensing strategy · Statutory damages across millions of articles create existential tail risk

Generated by claude-opus-5 on Aug 4, 2026, from this case’s record only. It stands until a major development changes the record. Not legal or investment advice. How predictions are made →

BACKGROUND

The New York Times sued OpenAI and Microsoft in December 2023, alleging that millions of Times articles were copied to train the GPT models behind ChatGPT and Copilot and that the products reproduce and closely mimic its journalism, including paywalled content. Judge Sidney H. Stein largely denied dismissal in March 2025, preserving the core direct and contributory infringement claims while dismissing DMCA theories, and in April 2025 the Judicial Panel on Multidistrict Litigation centralized the news-publisher and author actions before Judge Stein as In re OpenAI, Inc. Copyright Infringement Litigation (MDL No. 3143).

The MDL is in heavy discovery. In November 2025, Magistrate Judge Ona T. Wang ordered OpenAI to produce a sample of 20 million de-identified ChatGPT conversation logs, and Judge Stein affirmed that order on January 5, 2026 over OpenAI’s user-privacy objections. On July 9, 2026, the Times, the Daily News and fifteen other media plaintiffs moved for sanctions, alleging that OpenAI falsely told the court it could not search its training corpora and output logs for plaintiffs’ works while internally running exactly such searches, and that it deleted billions of ChatGPT outputs. They seek adverse findings, exclusion of the log sample and fees. Bellwether selection is expected in late 2026.

TIMELINE

JUL 9, 2026PROCEDURAL

OpenAI publicly denies sanctions allegations

Responding to the publishers’ sanctions motion, OpenAI spokesperson Drew Pusateri denied the allegations, saying the Times was making ‘blatantly false allegations’ as its case weakened, and that OpenAI would continue defending user privacy and fair use principles.

New York Times, publishers seek sanctions against OpenAI
JUL 9, 2026FILING

News plaintiffs move for sanctions against OpenAI

Seventeen media plaintiffs allege OpenAI concealed its ability to search training data and output logs and deleted billions of ChatGPT outputs. They seek adverse findings and fees.

New York Times says OpenAI hid evidence — TechCrunch
JAN 5, 2026PROCEDURAL

Judge Stein affirms the 20-million-log order

User-privacy objections are overruled. De-identification and the protective order are deemed sufficient safeguards.

When Chats Become Evidence — National Law Review
NOV 7, 2025PROCEDURAL

Magistrate Judge Wang orders production of 20 million ChatGPT logs

The court rejects OpenAI’s proposal to run targeted keyword searches and produce only conversations implicating plaintiffs’ works.

When Chats Become Evidence — National Law Review
APR 3, 2025PROCEDURAL

JPML centralizes the litigation as MDL No. 3143

Twelve news-publisher and author actions against OpenAI and Microsoft are consolidated before Judge Stein in the S.D.N.Y. More followed.

JPML Transfer Order
MAR 26, 2025PROCEDURAL

Judge Stein largely denies motions to dismiss

Core direct and contributory copyright claims proceed. DMCA claims are dismissed. The full opinion issued April 4, 2025.

S.D.N.Y. opinion, April 4, 2025
DEC 27, 2023FILING

New York Times sues OpenAI and Microsoft

The Times alleges millions of its articles were used without authorization to train GPT models and that ChatGPT reproduces its journalism.

The New York Times v. Microsoft and OpenAI — Wikipedia