Citation and evidence

New York Times v. OpenAI

23 min full readUpdated 24 references

This article's verification

Report a problem with this article

Edit article
More

Use this article

Raw MarkdownExplore connections

Improve this page

Revision historyDiscuss this article

Topics

AI Incidents & ControversiesAI Policy & RegulationOpenAI

Cite this article

The New York Times Company v. Microsoft Corporation, OpenAI Inc., et al. (case number 1:23-cv-11195) is a federal copyright lawsuit filed by The New York Times Company against Microsoft and OpenAI in the U.S. District Court for the Southern District of New York on December 27, 2023. The complaint alleges that OpenAI and Microsoft copied "millions" of Times articles to train large language models that power ChatGPT, Microsoft Copilot, and related products, and that those models reproduce Times content verbatim or near-verbatim and "hallucinate" false statements falsely attributed to the Times.[1][2] The Times seeks billions of dollars in statutory and actual damages, an injunction, and the destruction of any GPT or other language model and training set that incorporates Times works, a remedy that would be unprecedented in scope if granted.[1] It is widely considered the most consequential test of how United States copyright law applies to AI training and outputs, and a bellwether for whether fair use shields the training of generative models on copyrighted text.

The case is being heard by U.S. District Judge Sidney H. Stein, who in April 2025 denied most of OpenAI's and Microsoft's motions to dismiss and allowed direct copyright infringement, contributory infringement, certain Digital Millennium Copyright Act (DMCA) claims, and trademark dilution claims to proceed.[3] U.S. Magistrate Judge Ona T. Wang has presided over discovery, including a high-profile May 13, 2025 order requiring OpenAI to preserve all ChatGPT and API output logs that would otherwise have been deleted, an order Judge Stein later upheld.[4][5] The Times' case has been consolidated for pretrial purposes with parallel suits by a group of regional newspapers led by the New York Daily News and by the Center for Investigative Reporting (CIR).[3]

What does the New York Times allege?

By late 2023, the use of copyrighted text to train generative AI systems had become one of the most contested legal questions in the technology industry. OpenAI's GPT-3 (2020), GPT-3.5 (2022) and GPT-4 (2023) had been trained on enormous text corpora that included scraped web content from datasets such as Common Crawl, WebText, and Books2. OpenAI did not disclose the precise composition of its training data after GPT-2, but the Times' complaint, drawing on OpenAI's own published papers, alleged that Times content was heavily represented in those corpora, including Common Crawl.[1][6]

The November 2022 launch of ChatGPT, which reached an estimated 100 million users within three months of release, intensified scrutiny of how those models had been built.[6] Authors, news organizations, image platforms and individual artists began filing copyright suits against AI developers throughout 2023, including Sarah Silverman's class action against OpenAI and Meta, the Authors Guild action, and Getty Images' suit against Stability AI. The Times, which had separately objected to being scraped by web crawlers, blocked OpenAI's GPTBot from its websites in August 2023 (as early as August 17), shortly after OpenAI publicly documented the crawler.[7]

The Times had also engaged OpenAI directly. According to the Times' complaint, the Times reached out to Microsoft and OpenAI in April 2023 to raise intellectual property concerns and explore an amicable resolution, but "these efforts have not produced a resolution."[1] OpenAI's January 8, 2024 blog post "OpenAI and journalism" said the negotiations focused on "a high-value partnership around real-time display with attribution in ChatGPT" and that the discussions "had appeared to be progressing constructively through our last communication on December 19," eight days before the suit was filed.[8]

When was the complaint filed, and what does it claim?

The Times filed its 69-page complaint on December 27, 2023, captioned The New York Times Company v. Microsoft Corporation, OpenAI Inc., OpenAI LP, OpenAI GP LLC, OpenAI LLC, OpenAI OpCo LLC, OpenAI Global LLC, OAI Corporation LLC, and OpenAI Holdings LLC, case number 1:23-cv-11195 in the U.S. District Court for the Southern District of New York.[1] Lead trial counsel for the Times is Ian B. Crosby of Susman Godfrey, joined by lawyers from Rothwell, Figg, Ernst & Manbeck.[9]

Causes of action

The complaint asserted seven counts: (1) direct copyright infringement under 17 U.S.C. § 501; (2) vicarious copyright infringement; (3) contributory copyright infringement against Microsoft; (4) contributory copyright infringement against all defendants; (5) violations of the DMCA, 17 U.S.C. § 1202, for the removal or alteration of copyright management information (CMI); (6) common-law unfair competition by misappropriation ("hot news"); and (7) federal trademark dilution under 15 U.S.C. § 1125(c).[1] The Times later filed a First Amended Complaint, which is the pleading the April 2025 ruling addressed.[6]

Core allegations

The complaint frames the dispute in three principal stages.

First, the Times alleges that OpenAI and Microsoft made wholesale unauthorized copies of millions of Times articles when assembling training corpora and when training their GPT models. The complaint identifies Common Crawl, WebText, and WebText2 as datasets that, according to the Times, contained large volumes of scraped Times material (it says WebText contains "a staggering amount of scraped content from The Times"), and alleges that www.nytimes.com is the most highly represented proprietary source in a filtered English-language subset of Common Crawl.[1]

Second, the Times alleges that the GPT models themselves embody "unauthorized reproductions and derivatives" of Times works because their parameters encode copies of protected Times expression, and that those parameters allow the models to produce Times text on demand.[1]

Third, the Times alleges that ChatGPT and related products generate outputs that copy Times articles verbatim or near-verbatim. Exhibit J to the complaint presents one hundred examples in which a short prefix from a published Times article, supplied as a prompt to GPT-4, was followed by an output reproducing hundreds of additional words from the same article with only minor variation.[10] The complaint pairs these "memorization" examples with allegations that Bing search results powered by GPT-4 and the "Browse with Bing" plugin to ChatGPT generated "extensive paraphrases and direct quotes" of Times content without referring users back to the Times.[6]

A fourth allegation, given prominent treatment in the complaint, concerns hallucinations that misattribute fabricated content to the Times, including invented health and consumer-product recommendations purportedly drawn from Wirecutter, the Times' product-review site. The Times alleges that these fabrications harm its reputation, citing, for example, a GPT-4 answer that reproduced Wirecutter's top four office-chair picks but added two chairs Wirecutter never recommended and falsely attributed them to Wirecutter, and Bing Chat answers that invented content attributed to Times articles.[1] The Times argues that this use of its marks on "lower quality and inaccurate writing" dilutes its trademarks by tarnishment in violation of 15 U.S.C. § 1125(c).[1]

Relief sought

The Times asks for an order requiring "destruction under 17 U.S.C. § 503(b) of all GPT or other LLM models and training sets that incorporate Times Works," together with billions of dollars in statutory and actual damages.[1] Statutory damages under 17 U.S.C. § 504 can reach $150,000 per work for willful infringement, and the Times pleads that its registered, copyrighted works number in the millions.[6]

How did OpenAI and Microsoft respond?

On January 8, 2024, OpenAI published a blog post titled "OpenAI and journalism," that pushed back against the lawsuit on four fronts.[8] OpenAI stated that (1) it collaborates with news organizations and is "creating new opportunities" through partnerships and product features; (2) "Training AI models using publicly available internet materials is fair use, as supported by long-standing and widely accepted precedents"; (3) "'Regurgitation' is a rare bug that we are working to drive to zero"; and (4) the Times "is not telling the full story."[8] OpenAI asserted that "we had explained to The New York Times that, like any single source, their content didn't meaningfully contribute to the training of our existing models." Regarding the regurgitation examples, OpenAI suggested the Times "intentionally manipulated prompts, often including lengthy excerpts of articles, in order to get our model to regurgitate,".[8]

Microsoft filed its own motion to dismiss parts of the complaint in March 2024. It accused the Times of "doomsday futurology," said the Times had crafted "unrealistic prompts" to coax the GPT-based tools into reproducing its content, and compared the suit to Hollywood's effort to stop the VCR.[22]

OpenAI moved to dismiss several counts in February 2024, arguing that direct-infringement claims based on the creation and use of the GPT-2 and GPT-3 training datasets in 2019 and 2020 were time-barred under the Copyright Act's three-year statute of limitations in 17 U.S.C. § 507(b), and that the Times' DMCA, contributory infringement, and "hot news" misappropriation claims failed as a matter of law.[3] Microsoft moved separately on overlapping grounds. In the later Daily News action, Microsoft also challenged the New York state-law trademark dilution claim, while OpenAI challenged the federal dilution claim.[3][6]

What other lawsuits is the case consolidated with?

The Times' case was not the only newspaper suit against OpenAI and Microsoft in the Southern District of New York. On April 30, 2024, the Daily News LP and seven other regional newspapers, including the Chicago Tribune, the Orlando Sentinel, the Sun-Sentinel, the San Jose Mercury News, the Denver Post, the Orange County Register, and the Pioneer Press, filed a parallel complaint asserting essentially the same claims as the Times, with the addition of a New York General Business Law § 360-l trademark dilution claim. The case was docketed as 24-cv-3285.[3] In June 2024, the Center for Investigative Reporting, publisher of Mother Jones and Reveal, filed its own action against OpenAI and Microsoft, docketed as 24-cv-4872, with an amended complaint following on September 24, 2024.[3]

The three actions were consolidated for pretrial purposes before Judge Stein. On April 3, 2025, the Judicial Panel on Multidistrict Litigation centralized the copyright suits against OpenAI, including book-author class actions such as the Authors Guild case and suits transferred from California, before Judge Stein as In re OpenAI, Inc. Copyright Infringement Litigation (MDL No. 3143).[11][23] Throughout 2024 and 2025, Judge Stein's court became the principal forum for U.S. copyright disputes against OpenAI, with parallel disputes pending in California against Meta, Anthropic, and other developers.[11]

How did Judge Stein rule on the motion to dismiss (March-April 2025)?

On March 26, 2025, Judge Stein issued a brief order indicating how he would rule on the defendants' motions to dismiss, telling the parties an opinion would follow "expeditiously."[12] His full opinion was filed on April 4, 2025 and runs 43 pages. The opinion is a single ruling that decides motions in the Times, Daily News, and CIR actions together.[6]

Claims allowed to proceed

The Court denied OpenAI's and Microsoft's motions to dismiss the direct copyright infringement claims to the extent the defendants argued the claims were time-barred. Applying the Second Circuit's "discovery rule," Judge Stein held that the plaintiffs could pursue direct-infringement claims based on training conduct in 2019 and 2020 because they had plausibly alleged that they neither discovered nor should have discovered the alleged infringements before December 27, 2020 (for the Times) or April 30, 2021 (for the Daily News).[6] The opinion held that the few public documents from 2019 and 2020 that OpenAI cited did not show that the plaintiffs knew, or should have known, of the alleged infringement, and noted that there is no "sophisticated plaintiff" exception to the discovery rule.[6]

The Court also denied the defendants' motions to dismiss the contributory copyright infringement claims. Judge Stein concluded that the plaintiffs had plausibly alleged (i) third-party infringement by ChatGPT users, (ii) defendants' knowledge of such infringement (drawing on widely publicized regurgitation incidents and the more than 100 pages of examples in Exhibit J), and (iii) that defendants' tools were not protected by the "substantial noninfringing uses" doctrine at the pleading stage.[6]

In the Daily News action, the Court denied the defendants' motions to dismiss both the federal trademark dilution claim under 15 U.S.C. § 1125(c) and the New York General Business Law § 360-l trademark dilution claim, finding that the plaintiffs had adequately pleaded that their marks were "famous" and rejecting Microsoft's argument that the New York statute violates the Dormant Commerce Clause.[6]

The Court also denied OpenAI's motion to dismiss certain DMCA § 1202(b)(1) claims, namely those against OpenAI in the Daily News and CIR actions, allowing the plaintiffs to proceed on the theory that OpenAI intentionally removed copyright management information from copies of their works during training.[6]

Claims dismissed

Judge Stein granted the defendants' motions to dismiss the common-law unfair competition by misappropriation ("hot news") claims in all three actions, holding them preempted by the Copyright Act because the plaintiffs had not plausibly alleged the narrow "hot news" exception that traces to International News Service v. Associated Press (1918). These claims were dismissed with prejudice.[6]

The Court also granted OpenAI's motion to dismiss the "abridgment" claims in the CIR action, dismissed with prejudice on the ground that the abridgments alleged in the complaint were not substantially similar to CIR's copyrighted works as a matter of law.[6]

Several DMCA claims were dismissed without prejudice: the § 1202(b)(1) claims against Microsoft in all three actions, the § 1202(b)(1) claim against OpenAI in the Times' action specifically, and the § 1202(b)(3) claims against both defendants in all three actions.[6] The dismissal of the Times' § 1202(b)(1) claim against OpenAI was a notable setback for the Times in particular; Judge Stein found that the Times complaint gave no specific detail on how CMI was removed during training, and that because Exhibit J showed only excerpts it was "entirely plausible" that CMI remained on the training copies; the Daily News and CIR complaints, by contrast, described how the Dragnet and Newspaper extractors stripped CMI. Separately, on the § 1202(b)(3) claims, he reasoned that allowing DMCA claims where the output is not "close to identical" to the original would risk boundless DMCA liability.[6]

Significance of the ruling

The April 2025 opinion did not decide any of the merits questions that the case ultimately turns on, in particular whether OpenAI's training is protected by fair use. But by allowing the direct infringement, contributory infringement, federal and state trademark dilution, and certain DMCA claims to go forward, the ruling assured that the case would proceed to discovery and likely to a jury trial. The discovery-rule ruling mattered because it preserved direct-infringement claims over training conduct in 2019 and 2020, the period in which the GPT-2 and GPT-3 training datasets were created.[6][13]

Why was OpenAI ordered to preserve ChatGPT logs?

The most contested feature of the case in 2025 was a series of discovery disputes over the preservation and production of ChatGPT and API output logs. The Times argued that it needed access to actual ChatGPT outputs in order to test the prevalence of memorization and verbatim reproduction of its articles, and pressed for orders requiring OpenAI to preserve logs that the company would otherwise have deleted under its retention policies.

May 13, 2025 preservation order

On May 13, 2025, Magistrate Judge Ona T. Wang ordered OpenAI to "preserve and segregate all output log data that would otherwise be deleted on a going forward basis until further order of the Court."[4] The order applied to consumer ChatGPT logs (including for Free, Plus, Pro, and Team users) and to API outputs that OpenAI was not contractually required to retain. It overrode OpenAI's practice of permanently deleting deleted chats within 30 days. According to OpenAI, it did not apply to ChatGPT Enterprise or ChatGPT Edu customers, or to API customers using Zero Data Retention endpoints.[20]

OpenAI publicly criticized the order. In a June 5, 2025 post, OpenAI COO Brad Lightcap called the Times' demand "sweeping and unnecessary" and said it "abandons long-standing privacy norms and weakens privacy protections," adding that OpenAI would keep appealing what it called an "overreach."[20] Magistrate Judge Wang had already denied OpenAI's motion for reconsideration on May 16, 2025, without prejudice to renewal, and set a May 27 conference on the alleged spoliation and the proportionality of continued preservation.

June-July 2025: District Court affirmance

OpenAI appealed Magistrate Judge Wang's preservation order to Judge Stein. After oral argument, Judge Stein upheld the preservation order in late June 2025, prioritizing the litigation hold over OpenAI's user-privacy arguments.[15] In July 2025, the Times moved to compel production of a sample of 120 million chat logs from December 2022 through November 2024 in order to study the frequency of Times-related outputs.

November 7, 2025 production order

After months of negotiation between the parties, OpenAI had offered the Times a 20-million-log sample (about 0.5% of its preserved logs), which the plaintiffs accepted, but OpenAI then changed position in October 2025, proposing to run keyword searches over the preserved logs and produce only conversations that implicated specific plaintiffs' works. On November 7, 2025, Magistrate Judge Wang rejected that proposal and ordered OpenAI to produce the full 20-million-log de-identified sample, subject to "multiple layers of protection" for user privacy including de-identification, an existing protective order, and an "attorneys' eyes only" designation.[16]

January 2026 affirmance

OpenAI again appealed to Judge Stein. On January 5, 2026, Judge Stein affirmed Magistrate Judge Wang's order, holding that "Judge Wang's rulings were neither clearly erroneous nor contrary to law" and that user privacy interests were sufficiently protected by the de-identification process and the protective order already in place.[16][15] Judge Stein distinguished a securities-law precedent on which OpenAI had relied involving SEC wiretap recordings, noting that ChatGPT users had voluntarily submitted their communications and that OpenAI's ownership of the logs was uncontested. The opinion rejected OpenAI's argument that the law required the court to "order the least burdensome discovery possible."[15]

The preservation aspect of the order ended on September 26, 2025, allowing OpenAI to resume normal data-retention practices for newly generated content; logs already preserved under the order remained available for the litigation, and OpenAI had to keep preserving data for ChatGPT accounts flagged by the Times; the change came in an October 9, 2025 order by Magistrate Judge Wang.[21] The dispute spurred substantial public debate over the privacy implications of broad discovery in AI litigation.

While the Times case has been the most visible, by 2026 it sat within a broader ecosystem of AI-training copyright disputes that produced several mixed signals on the central question of fair use. In June 2025, in two separate cases against Anthropic and Meta involving books, two California federal judges held that training large language models on copyrighted books was sufficiently transformative to constitute fair use, while reaching different conclusions on liability for the way training corpora were assembled.[17] In Bartz v. Anthropic, Judge William Alsup ruled that Anthropic's downloading and retention of more than 7 million pirated books to assemble a permanent training library was not fair use, even though training itself was "quintessentially transformative"; the case later settled for approximately $1.5 billion in September 2025, in what was widely reported as the largest copyright settlement in U.S. history.[17] In the parallel Kadrey v. Meta action, Judge Vince Chhabria reached a defense-favorable result on the specific record before him while warning that the outcome could differ in other cases.[17]

OpenAI has cited those fair-use rulings in its filings against the Times, arguing that the same reasoning should govern. The Times has responded that its case differs in several respects, most notably the extensive evidence of verbatim memorization of its articles and the use of GPT-powered tools (Bing, Microsoft Copilot, Browse with Bing) as competitive substitutes for the Times itself.[14]

The Times' case is widely seen as the highest-stakes copyright dispute of the generative-AI era because of the size of the works at issue (more than 10 million registered, copyrighted works by the time of the First Amended Complaint, according to the Times),[6] the directly competitive relationship between OpenAI's products and the Times' subscription journalism, and the scope of relief the Times has demanded, including the destruction of model weights.

Current status (as of November 2026)

The case narrowed in 2026. After the Supreme Court held in Cox Communications, Inc. v. Sony Music Entertainment (March 25, 2026) that "material contribution" is not a valid theory of contributory copyright liability, the Times, the Daily News plaintiffs and CIR dropped their material-contribution claims based on infringement by ChatGPT and Copilot users, and the Times and the Daily News dropped their trademark-dilution claims. Judge Stein dismissed those claims with prejudice in an August 6, 2026 memorandum order and refused to let the Times add new secondary-liability theories against Microsoft. The direct-infringement claims were not affected.[23] On July 9, 2026, the Times and other news plaintiffs moved for sanctions, alleging that OpenAI had told the court it could not search its training data and ChatGPT logs for their articles when it could, and that it had failed to preserve conversations it was ordered to keep; the motion was fully briefed in September 2026.[23] On September 4, 2026, the news plaintiffs, OpenAI and Microsoft filed cross-motions for summary judgment, with OpenAI asking the court to rule that pretraining on the articles is fair use; the U.S. Justice Department had filed a statement of interest supporting fair use for AI training on September 1, and the parties filed oppositions on October 23, 2026.[23] As of November 2026, the court had not ruled on summary judgment or sanctions, no trial date had been set, and no settlement had been reached.[23]

In May 2026, attention in the AI industry was largely focused on the Northern District of California trial in Musk v. Altman, which began jury selection on April 27, 2026 and concluded closing arguments on May 14, 2026.[19] That case, while addressing different legal questions (governance and contract claims rather than copyright), reinforced the view that 2026 would be a pivotal year for the courts' treatment of OpenAI and the artificial-intelligence sector as a whole.

The New York Times v. OpenAI case has already had practical consequences regardless of its ultimate outcome.

First, the case produced one of the first major decisions allowing newspaper copyright claims against an AI developer to proceed past the pleading stage. Its contributory-infringement holding did not last: after the Supreme Court's March 2026 decision in Cox Communications v. Sony Music Entertainment, the Times dropped those claims and Judge Stein dismissed them with prejudice in August 2026.[23]

Second, the preservation and production orders concerning ChatGPT logs raised novel issues at the intersection of e-discovery, user privacy, and AI product design. Companies that operate consumer LLM products have begun reassessing their data-retention policies in light of the possibility of similar orders in other litigation. In the Times' case a court ordered open-ended preservation of consumer LLM conversation logs (lifted for new data as of September 26, 2025) and ultimately the production of a 20-million-log sample.[16][21]

Third, the case has spurred parallel licensing deals between AI developers and news publishers. OpenAI signed content-licensing agreements with publishers both shortly before and after the Times' filing, including the Associated Press, Axel Springer (publisher of Politico and Business Insider), the Financial Times, News Corp (publisher of The Wall Street Journal), Vox Media, Dotdash Meredith, and others. The Times itself signed its first generative-AI licensing deal in May 2025, with Amazon, covering content from the Times, NYT Cooking and The Athletic.[24]

Fourth, the case has reshaped the public debate on whether the training of generative-AI systems on copyrighted works is permissible under existing fair-use doctrine, on whether trained model weights can themselves be infringing derivative works, and on the appropriate remedy if liability is found. The remedy question is particularly stark in the Times' case because the company has demanded the destruction of GPT models that incorporate Times works, a remedy that would be unprecedented in scope if granted.[1]

Whatever its ultimate disposition, the Times' case is likely to remain a benchmark for U.S. copyright law's response to generative AI for years to come.

Continue reading

More to read

Topics:AI Incidents & ControversiesAI Policy & RegulationOpenAI

References

  1. ^1 ^2 ^3 ^4 ^5 ^6 ^7 ^8 ^9 ^10 ^11 ^12The New York Times Company, "Complaint in The New York Times Company v. Microsoft Corporation et al.," U.S. District Court for the Southern District of New York, Case No. 1:23-cv-11195, December 27, 2023. nytco-assets.nytimes.com/..._Complaint_Dec2023.pdf. Accessed 2026-05-19.
  2. ^Michael M. Grynbaum and Ryan Mac, "The Times Sues OpenAI and Microsoft Over A.I. Use of Copyrighted Work," The New York Times, December 27, 2023. nytimes.com/...ork-times-open-ai-microsoft-lawsuit. Accessed 2026-05-19.
  3. ^1 ^2 ^3 ^4 ^5 ^6Loeb & Loeb LLP, "New York Times v. Microsoft Corp.," April 2025. loeb.com/...new-york-times-v-microsoft-corp. Accessed 2026-05-19.
  4. ^1 ^2Justia Dockets, "ORDER: Accordingly, OpenAI is NOW DIRECTED to preserve and segregate all output log data in The New York Times Company v. Microsoft Corporation et al.," Docket No. 551, May 13, 2025. docs.justia.com/...551. Accessed 2026-05-19.
  5. ^Thurrott.com, "OpenAI Must Turn Over ChatGPT Logs in New York Times Case," 2025. thurrott.com/...hatgpt-logs-in-new-york-times-case. Accessed 2026-05-19.
  6. ^1 ^2 ^3 ^4 ^5 ^6 ^7 ^8 ^9 ^10 ^11 ^12 ^13 ^14 ^15 ^16 ^17 ^18Sidney H. Stein, U.S. District Judge, "Opinion" in The New York Times Company v. Microsoft Corporation et al., No. 23-cv-11195 (SHS), Daily News LP, et al. v. Microsoft Corporation et al., No. 24-cv-3285 (SHS), and The Center for Investigative Reporting, Inc. v. OpenAI, Inc. et al., No. 24-cv-4872 (SHS), U.S. District Court for the Southern District of New York, April 4, 2025. nysd.uscourts.gov/...pinion%20april%204%202025.pdf. Accessed 2026-05-19.
  7. ^Blake Brittain, "NY Times sues OpenAI, Microsoft for infringing copyrighted works," Reuters, December 27, 2023. reuters.com/...ringing-copyrighted-work-2023-12-27. Accessed 2026-05-19.
  8. ^1 ^2 ^3 ^4OpenAI, "OpenAI and journalism," OpenAI blog, January 8, 2024. openai.com/...openai-and-journalism. Accessed 2026-05-19.
  9. ^Susman Godfrey LLP, "Ian B. Crosby." susmangodfrey.com/...ian-b-crosby. Accessed 2026-05-19.
  10. ^ChatGPT Is Eating the World, "Exhibit J to New York Times Complaint provides 'One Hundred Examples of GPT-4 Memorizing Content from the New York Times,'" December 27, 2023. chatgptiseatingtheworld.com/...-the-new-york-times. Accessed 2026-05-19.
  11. ^1 ^2McKool Smith, "AI Infringement Case Updates: November 24, 2025." mckoolsmith.com/newsroom-ailitigation-46. Accessed 2026-05-19.
  12. ^NPR (Bobby Allyn), "Judge allows 'New York Times' copyright case against OpenAI to go forward," March 26, 2025. npr.org/...imes-openai-copyright-case-goes-forward. Accessed 2026-05-19.
  13. ^ChatGPT Is Eating the World, "Judge Stein issues opinion rejecting most of OpenAI's motion to dismiss claims by newspapers, book authors," April 5, 2025. chatgptiseatingtheworld.com/...-v-openai-microsoft. Accessed 2026-05-19.
  14. ^OpenAI, "Reporting the facts about The New York Times' lawsuit." openai.com/new-york-times. Accessed 2026-05-19.
  15. ^1 ^2 ^3Bloomberg Law (Annelise Levy), "OpenAI Must Turn Over 20 Million ChatGPT Logs, Judge Affirms," January 2026. news.bloomberglaw.com/...hatgpt-logs-judge-affirms. Accessed 2026-05-19.
  16. ^1 ^2 ^3ABA Journal, "ChatGPT creator must turn over 20M chat logs in copyright litigation, federal judge says," January 2026. abajournal.com/...ht-litigation-federal-judge-says. Accessed 2026-05-19.
  17. ^1 ^2 ^3Public Knowledge, "Courts Agree: AI Training Ruled As Fair Use in Bartz v. Anthropic and Kadrey v. Meta," 2025. publicknowledge.org/...anthropic-and-kadrey-v-meta. Accessed 2026-05-19.
  18. Originality.AI, "OpenAI and ChatGPT Lawsuit List." originality.ai/...openai-chatgpt-lawsuit-list. Accessed 2026-05-19.
  19. ^CNBC, "Closing arguments conclude in Musk v. Altman, jury to deliberate next week," May 14, 2026. cnbc.com/...sing-arguments-jury-openai-musk-altman. Accessed 2026-05-19.
  20. ^1 ^2OpenAI (Brad Lightcap), "How we're responding to The New York Times' data demands in order to protect user privacy," June 5, 2025. openai.com/...response-to-nyt-data-demands. Archived June 2025 copy: web.archive.org/...response-to-nyt-data-demands. Accessed 2026-11-05.
  21. ^1 ^2Jackson Chen, "OpenAI no longer has to preserve all of its ChatGPT data, with some exceptions," Engadget, October 11, 2025. engadget.com/...ata-with-some-exceptions-192422093. Accessed 2026-11-05.
  22. ^Mariella Moon, "Microsoft accuses the New York Times of doom-mongering in OpenAI lawsuit," Engadget, March 5, 2024. engadget.com/...gering-in-openai-lawsuit-133025748. Accessed 2026-11-05.
  23. ^1 ^2 ^3 ^4 ^5 ^6Steve Levine, "New York Times OpenAI Lawsuit: Where the Case Stands," OpenClassActions.com, October 6, 2026 (updated October 27, 2026). openclassactions.com/...icrosoft-copyright-lawsuit. Accessed 2026-11-05.
  24. ^Rebecca Bellan, "The New York Times and Amazon ink AI licensing deal," TechCrunch, May 29, 2025. techcrunch.com/...and-amazon-ink-ai-licensing-deal. Accessed 2026-11-05.

Improve this article

Add missing citations, update stale details, or write a clearer explanation. Anyone can edit: changes go live when you publish, and every edit is checked by an editor.

6 revisions by 1 contributor · v6 · 4,547 words · full history

Fact-checks are independent of edits: a reviewer re-verifies the article against its sources and stamps the date. How we verify

Research and drafting on this wiki are AI-assisted, under named human editorial standards. How AI is used here

Reviewer note: Independent fact-check of the full article against its cited sources (130 claims, 27 sources fetched); 12 material + 18 minor fixed

Cite this page: AI Wiki. "New York Times v. OpenAI." aiwiki.ai, updated 24 Jun 2026, fact-checked 5 Nov 2026. CC BY 4.0. https://aiwiki.ai/wiki/new_york_times_v_openai

Edit article