OpenAI on Friday, October 30, 2026 published excerpts from 20 court filings by authors, librarians, law professors, trade groups, a venture capital firm, two former members of Congress and the United States government, all backing fair use for AI training, and accused The New York Times of pursuing “an undeserved payday” in the copyright case the newspaper filed against it in December 2023.

The post, titled “Fair use benefits everyone,” appeared in the Company section of OpenAI’s website, credited to the company rather than to a named author. OpenAI has also rewritten the opening of its long-running page on the litigation, “Reporting the facts about the New York Times’ lawsuit,” which now leads with the new post and adds 2026 entries on a granted motion to strike, the summary judgment motions and two orders narrowing claims. Both documents are statements by a party to active litigation. They set out OpenAI’s position, not findings of the court.

In Short

The New York Times sued OpenAI, saying the company copied millions of its articles without permission to build ChatGPT. OpenAI says using articles to teach an AI system is allowed under a long-standing rule called fair use, and on October 30 it published quotes from 20 groups and people, including the US government, who told the judge they agree. The Times and the other news plaintiffs argue the copying is not fair use and ask for damages per article. A judge in New York still has to decide, and the answer will shape whether AI companies must pay publishers to use their work.

Twenty excerpts, one target

The October 30 post opens with a broad claim. “For decades, fair use has played a critical role in driving innovation that benefits everyone,” according to OpenAI, which says the Times’ lawsuit “asks everyone to ignore a half century of legal history and progress, and toss aside the longstanding principle of fair use in their own narrow pursuit of an undeserved payday.”

What follows is a selection. OpenAI describes the material as “excerpts from some of those submissions, in their own words,” and says the full filings are available on the public court docket. The post does not say how many briefs were filed in total, and it does not mention any filed in support of the news plaintiffs. Nor does it include the Wikimedia Foundation, which asked the court on October 29, 2026 for leave to file a brief in support of neither party. That filing agrees that training the models behind ChatGPT was likely fair use, but argues that the chatbot’s practice of fetching web pages and reproducing them in its answers likely is not.

The 20 entries span an unusual range. Filers on the list include AI Progress; a joint filing by the American Council of the Blind, the Association of Research Libraries, the American Library Association, the Consumer Technology Association and the Business Software Alliance; Authors Alliance; Chamber of Progress with the Abundance Institute and Trevor Lawrence Jr.; the Computer and Communications Industry Association; the Electronic Frontier Foundation; Engine with the Developers Alliance; the Foundation for American Innovation; and Public Knowledge. Academics include law professors Ed Lee and Rebecca Tushnet, Matthew Sag, and Brian L. Frye, who filed with Pulitzer Prize-winning journalist Ron Suskind, as well as Joshua S. Gans, a professor of strategic management. Computer scientists Mike Knoop, Daniel Lowd and Douglas Schmidt filed together. Author Kevin Kelly, library technology expert Gregory Leppert, musician and composer Michael Wall and comedian and writer Sarah Rose Siskind appear as individuals. Khosla Ventures filed as an investor. Former Senator Rick Santorum and former Representative Rick Boucher filed jointly. The last entry is the United States Government.

Read together, the excerpts make four arguments.

What the briefs argueTraining as analysis, not copying

The first line of argument is about transformation, the first of the four fair use factors, and it turns on how large language models are built. “[T]raining a generative AI model is among the most clearly transformative uses a court has been asked to evaluate,” according to the Computer and Communications Industry Association excerpt. “The copyrighted works are not copied to be consumed or distributed; they are analyzed to extract statistical patterns that enable a tool to generate new text or other output at the direction of a human user.”

The computer scientists took the technical route. Generalization and memorization are distinct, according to Knoop, Lowd and Schmidt, and “A trained model is not an addressable document repository and does not ordinarily retrieve identified training data as stored files.” There is, they write, “no address, index, or file handle by which particular training data can be requested and returned; generation instead proceeds by producing a probability distribution for the next token and selecting from it.”

Santorum and Boucher went further. “The use of copyrighted works to train AI models is a paradigmatic case of fair use,” the former lawmakers wrote, citing the Supreme Court’s 1984 Sony decision on the doctrine’s relationship with new technology. Sag argued that “under realistic conditions the outputs of defendants’ models are in no relevant way similar to the invisible intermediate copies made during the training process.” Engine and the Developers Alliance tied the point to startups that build on foundation models: “The purpose of training a foundation model is not to reproduce the works it learns; it is to build a general-purpose tool that others can adapt to new ends.”

That claim has been tested outside the courtroom. In January 2026, Stanford researchers reported extracting long passages of copyrighted books from production models, reproducing 95.8% of one novel from Claude 3.7 Sonnet after jailbreak attempts and 76.8% from Gemini 2.5 Pro without one, while GPT-4.1 stopped near 4% after 5,179 attempts. The results varied sharply by model, which is part of why the record in each case matters.

The fight over market dilution

The second line of argument targets a specific theory: market dilution, the idea that AI outputs harm authors by flooding the market with competing works, even when those outputs do not copy protected expression. The theory goes to the fourth fair use factor, the effect of the use on the market for the original.

Authors Alliance asked the court to hold that “market dilution,” which it defines as “aggregate competitive pressure from noninfringing outputs, including those that occupy the same subject matter or genre, that share unprotected characteristics, or compete for the same audience, is not cognizable market harm under 17 U.S.C. § 107(4).” Its conclusion: “copyright protects authors against substitution for their protected expression, not against the success of new expression that merely competes with it.”

Lee and Tushnet were blunter. “Plaintiff’s unprecedented theory of ‘dilution’ turns copyright on its head,” the professors wrote. “Genres and facts are uncopyrightable.” Gans framed it in economic terms, arguing that copyright supports both the incentive to create content and the incentive to build tools that learn from it, and that “a rule that treated every displaced sale as actionable can weaken the second without reliably protecting the first.”

Suskind and Frye addressed the news plaintiffs directly. The papers “commissioned and published newspaper articles with the expectation of selling consumers access to those articles,” according to their excerpt, and “Their ability to sell consumers access to their articles wasn’t affected by OpenAI’s use of those articles to train its AI models.” The Electronic Frontier Foundation took on a second theory, saying the plaintiffs want the court “to rewrite copyright law to include a lawful acquisition requirement that finds no basis in statute or case law.”

The dilution theory is not a fringe position, however. The US Copyright Office’s pre-publication report on generative AI training, released in May 2025, said training foundation models will often be transformative but also said the fourth factor analysis should not ignore market dilution. The amicus excerpts OpenAI selected do not mention that report.

Access, libraries and the national interest

A third group of excerpts argues that the case reaches far beyond AI developers. The joint filing led by the American Council of the Blind says the case “affects medical researchers and patients, educators serving students with different learning needs, people with disabilities seeking greater independence, farmers applying agricultural knowledge, independent creators developing new forms of expression, and industry sectors from software to the law.” Many of those users “depend on systems developed by others, including open source AI systems,” it adds, and “Restrictions on fair use training would affect both the institutions building AI and the people using it.”

Public Knowledge placed AI training in a line of fair use precedents: image search in Perfect 10 v. Amazon (Ninth Circuit, 2007), full-text book search in Authors Guild v. Google (Second Circuit, 2015), accessible books for readers with print disabilities in Authors Guild v. HathiTrust (Second Circuit, 2014), and free publication of standards incorporated into law in ASTM v. Public.Resource.Org (D.C. Circuit, 2023). The Second Circuit cases carry particular weight, because the Southern District of New York sits within that circuit.

Leppert quoted the missions of the Library of Congress, Harvard Library, MIT Libraries and the New York Public Library to argue that openly available general-purpose models help libraries preserve and provide access to knowledge. The Foundation for American Innovation proposed what it calls “a novel application” of the fourth factor: “considering the national-security value of capabilities that enable new analysis, software, and other productive work.” Khosla Ventures asked the court to hold that training OpenAI’s models “constitutes fair use as a matter of law,” and AI Progress said “Settled fair use principles therefore compel the conclusion that generative AI training is fair use.”

Creators who use the tools

The fourth line puts working artists forward. Chamber of Progress, the Abundance Institute and Trevor Lawrence Jr. wrote that “musicians use generative AI to explore arrangements, develop melodies, and bring projects within reach that limited resources might otherwise foreclose.”

Wall and Siskind described their own businesses. Siskind uses AI “in every aspect of her business, from day-to-day administrative tasks to developing ideas and bringing in new clients,” according to their brief, which quotes her: “I don’t think I could run my business without AI at this point.” Wall, it says, doubts his music business “could survive in its current form without AI.” Kevin Kelly, the author, argued that “each AI model can be understood as an entirely new medium that offers authors the capability to create new texts and ideas, and to reason,” and that a ruling for the plaintiffs “would hinder the development of technologies capable of disseminating previously unimaginable quantities of information.”

The federal government’s position

The entry carrying the most institutional weight sits at the bottom of the list. The Justice Department’s statement of interest in the consolidated case, dated September 1, 2026, argues that training on copyrighted text is fair use. In the excerpt OpenAI chose, the government says “the New York Times seeks to narrow fair-use doctrine to exclude the training of OpenAI’s large language models (LLMs),” a result it calls “inconsistent with basic copyright law principles.”

The government’s filing also makes an argument that OpenAI repeats on its own case page: that “publications – including independent and start-up media, in addition to the New York Times itself – are leveraging LLMs for journalistic and other purposes.” Constraining LLM development “under a misunderstanding of fair use doctrine would thwart such creative and scientific progress while hindering American prosperity and economic mobility,” according to the excerpt.

A rewritten case page

OpenAI maintains its page on the lawsuit as a running record, adding entries as the case moves. The current version opens with a new framing. “The New York Times has tried to present this lawsuit as a binary choice between human authorship and AI. That’s simply wrong,” according to OpenAI. “This case is not about human authorship versus AI. It’s about the New York Times looking for an undeserved payday at the expense of progress that benefits everyone.”

The page makes several factual claims along the way. OpenAI says “Over a billion people each week” use its models. In November 2025, OpenAI’s chief information security officer put ChatGPT’s weekly users at 800 million. It says the Times used OpenAI’s technology “in service of a Pulitzer-winning story,” citing an August 2026 Nieman Lab report on Pulitzer winners that disclosed AI use. And it argues that “The New York Times’ business is healthier than ever.”

That last claim has numbers behind it. The Times reported second-quarter 2026 results on August 5, with digital advertising revenue up 20.7% to $114.0 million, total revenue up 11.2% to $762.5 million and about 13.35 million subscribers. The same results also showed the cost of the litigation: $4.6 million in pre-tax costs in the quarter for lawsuits over the unauthorized copying of its journalism by generative AI companies, up from $3.5 million a year earlier, and $8.8 million in the first half. Chief executive Meredith Kopit Levien said on the earnings call that the company’s aim is to become less reliant on intermediaries.

OpenAI also says it is “sensitive to the financial challenges that have faced others in the news industry over the past two decades, even though those concerns predated ChatGPT,” pointing to Northwestern University’s 2022 State of Local News report as an example. It says it “led the AI industry by pioneering ways for organizations to express their preferences regarding how OpenAI’s crawlers access their content to appear in search results or train foundation models,” a reference to controls such as robots.txt directives. On October 30, the same day as the fair use post, OpenAI published a separate post titled “Making it easier for websites to express AI training preferences.”

The page lists more than 20 publications that have partnered with OpenAI, including Axios, The Atlantic, Condé Nast, Dotdash Meredith, Hearst, News Corp, Prisa Media and Vox Media, and names earlier partners such as the Associated Press, Axel Springer, the American Journalism Project and New York University. It also describes a September 2026 program with WAN-IFRA and the Association of Independent Regional Press Publishers of Ukraine. Oksana Brovko, AIRPPU’s chief executive, is quoted as saying that the program “should give journalists back time to do journalism.”

OpenAI’s account of the 2026 docket

The 2026 entries cover three court events and the October 30 post.

On August 6, 2026, the court issued orders in the Times case and in the Ziff Davis case that OpenAI says “once again dismissed claims,” adding that both cases “continue proving not to hold up under scrutiny.” The page does not specify which claims were dismissed. Ziff Davis, owner of CNET, PCMag, IGN and Mashable, filed its suit in April 2025 in Delaware; the case was later consolidated with the Times litigation in New York.

On September 4, 2026, OpenAI filed summary judgment memoranda in the Authors Guild case and in the Times case. The motions were part of a broader round. OpenAI, Microsoft and five groups of news plaintiffs each moved for summary judgment that day before Judge Sidney H. Stein in multidistrict litigation 25-md-3143, with Magistrate Judge Ona T. Wang handling discovery. OpenAI’s brief counted more than 10.8 million asserted works, 6,030,928 of them Times articles dated from 1928 to 2024. OpenAI’s experts found 24 verbatim outputs in 20 million produced chat logs, or 0.00012%, with the longest passages at 29 and 43 words. The company also argued that more than 96% of the claimed infringement by its browsing feature involved only Bing snippets, and that copies made before publishers blocked its ChatGPT-User agent were impliedly licensed. OpenAI disclosed that agent in March 2023, while publishers did not block it until April 2024, according to the motion. Its brief says the plaintiffs’ own testing, with more than 250 million prompts, produced partial regurgitation of under 2.6% of works, and compares that with the 16% figure in the Google Books case, where the Second Circuit found fair use.

The plaintiffs see a different record. Their 92-page brief, signed by Ian Crosby of Susman Godfrey, sought judgment on liability for copying at five stages, asked the court to reject fair use and requested statutory damages per article rather than per issue. Their experts said 6.3 million works were used in training. Both defendants asked for oral argument.

The third entry is new. On September 23, 2026, OpenAI filed a motion “pointing out that plaintiffs’ counsel failed to disclose that they had funded one of the key studies they cited,” according to the page. “The court granted our motion and plaintiffs withdrew the study,” OpenAI writes, citing an order dated October 6, 2026. The page does not identify the study, what it measured or the court’s reasoning.

The privacy dispute over 20 million conversations

The longest section of the page concerns user data. OpenAI says the Times initially demanded 1.4 billion private ChatGPT conversations in May 2025, and earlier asked the court to require OpenAI to retain all user conversations indefinitely, including chats users had deleted. OpenAI says it fought the retention demand “and restored our users’ control over their private chats.”

The narrower demand did not go away. On November 7, 2025, Magistrate Judge Wang ordered OpenAI to produce 20 million de-identified consumer ChatGPT logs, a random sample of complete conversations from December 2022 to November 2024 that excluded Enterprise, Edu, Business and API customers. Wang denied OpenAI’s request for a stay on November 13, 2025. A Times spokesperson said at the time that the court had ordered a sample of chats anonymized by OpenAI itself under a protective order, and that no user’s privacy was at risk.

OpenAI’s account adds what happened next. An update dated December 16, 2025 says the company was obligated to comply while it appealed to the district judge, and that the data had “undergone a de-identification process intended to remove or mask PII and other private information.” It is being provided, according to OpenAI, “under tight access controls designed to prevent the Times from copying and printing data that isn’t directly relevant to this case.” A later, undated update says the district judge has ruled on the appeal and that OpenAI has complied with the order. The page does not describe the de-identification method.

The page also takes aim at a competitor’s choice. “Both the Magistrate Judge and the Times relied on the fact that Anthropic was willing to produce five million of their users’ conversations in litigation irrespective of relevance,” according to OpenAI. “However, Anthropic’s decision does not establish a precedent here, and does not justify invading the privacy of our users.” Wang’s November 2025 order did rely on Concord Music Group v. Anthropic, a California case in which a 5 million-record sample was ordered produced.

Those logs later became evidence. The 24 verbatim outputs OpenAI’s experts reported in the summary judgment round were found in the same 20 million logs.

Discovery fights and disclosure lines

Older entries on the page recount discovery disputes, all from OpenAI’s perspective. In November 2024, the Times said OpenAI had “destroyed” data. OpenAI’s version is that the Times requested a setup change on a review machine, and that the change “wiped the folder structure on a temporary cache drive they were incorrectly using to store their searches.” In OpenAI’s words: “No actual data was lost. All the Times needed to do was re-run their searches.” In a July 29, 2025 letter, OpenAI said the Times had “secretly deleted evidence of their extensive use of OpenAI’s models internally,” which the page calls “a fact that was undisputed in court.”

Discovery has not run only one way. In December 2024, Wang ordered OpenAI to produce employees’ work-related social media messages in the consolidated Authors Guild and Times cases, rejecting OpenAI’s argument that a California labor statute barred access.

The page also criticizes the Times’ journalism. It reproduces the parenthetical lines the newspaper inserts in stories about OpenAI, such as “(The New York Times has sued OpenAI and Microsoft, claiming copyright infringement of news content related to A.I. systems. The companies have denied those claims.)” and says: “While disclosures are common, giving them this level of prominence – often by inserting in the middle of stories – is not.” It adds that the Times “largely avoided informing its readers” about rulings favorable to AI developers.

OpenAI’s first legal response, a February 2024 motion to dismiss, argued that the Times “went to elaborate lengths to intentionally manipulate prompts in order to manufacture forced regurgitation.” In a related case, a judge dismissed DMCA claims brought by Raw Story Media and AlterNet in November 2024, finding the alleged harm from removal of copyright information too speculative, according to OpenAI. On December 15, 2025, the court dismissed a number of Ziff Davis’s claims, including its trademark and DMCA allegations, OpenAI says.

Bartz, Kadrey and what the page leaves out

OpenAI’s legal argument leans on two June 2025 rulings from the Northern District of California. “Two federal judges in two separate cases have already independently confirmed what copyright law has long supported, finding in those cases that training AI models is highly transformative and protected by fair use,” according to the page. It quotes The Conference Board: “These cases are significant victories for developers… In both cases, the judges noted that each AI model’s use of the copyrighted material was highly transformative, a key element of satisfying the fair use doctrine.”

The page is accurate about the training holdings. It leaves out how the cases ended. In Bartz v. Anthropic, Judge William Alsup held on June 23, 2025 that training on books was fair use, but rejected fair use for Anthropic’s central library of pirated copies. That finding exposed Anthropic to statutory damages and led to the $1.5 billion settlement filed in September 2025, covering roughly 500,000 works at about $3,000 per work. In Kadrey v. Meta, Judge Vince Chhabria granted Meta summary judgment on June 25, 2025, but the ruling applied only to the 13 authors who sued, and the court found that Llama produced no more than 50 words from any plaintiff’s book. Neither district court decision binds Judge Stein in New York.

The page also quotes Jason Kwon, OpenAI’s chief strategy officer: “AI models are profoundly transformative. They use massive computational power to learn deep mathematical patterns, analyses, and insights from trillions of datapoints so they can create new content – or even ‘think’ in an inner monologue like a person.”

OpenAI closes the page with the statute’s four non-exhaustive factors – the purpose and character of the use, the nature of the copyrighted work, the amount and substantiality of the portion used, and the effect on the potential market – and its own conclusion: “We believe generative AI training and use qualify as fair use at a minimum because they are transformative, non-expressive analytical uses that do not substitute for, or harm, the markets for the original works.” The sentence claims both training and use, a broader position than most of the amicus excerpts, which concentrate on training. As evidence of transformative use, OpenAI cites science and drug development work, a project with the Government of Iceland to help preserve the Icelandic language, its December 2025 state of enterprise AI report and GPT-5.2’s state-of-the-art result in late 2025 on GDPval, an evaluation covering knowledge-work tasks across 44 professions.

What the Times says

The Times’ public case has come mostly through its court filings and its chief executive. Kopit Levien set out the company’s position in an August 2026 interview with Bloomberg’s Odd Lots podcast. The Times spent close to $2 billion last year producing about half a million works, she said, and any AI license has to meet three tests: continuity with the publisher’s commercial trajectory, permission with control attached, and a sustainable and fair value exchange. The company has sued OpenAI and Microsoft, and separately Perplexity, while signing a licensing agreement with Amazon. Kopit Levien described enforcement in court as one way to address the problem and said the Times is open to deals when the terms can be found.

That last point complicates the “undeserved payday” framing. The Times does license its work for AI use; the dispute is over whether OpenAI needed a license at all. Is the case about payment or permission? On the Times’ account, it is both.

Why this matters for publishers and marketers

For publishers, the fair use question sits underneath every AI licensing negotiation. If training on news articles is fair use, the deals OpenAI lists on its page, from Axios to News Corp, are commercial choices rather than legal necessities, and a publisher’s leverage rests more on access controls and the value of current reporting than on copyright in its archive. If the court rejects fair use, those agreements look less like optional partnerships and more like the template for a market AI developers would have to enter, with the statutory damages the plaintiffs seek, calculated per article across millions of works, as the alternative.

The Wikimedia filing points to a second front. Even if training is held to be fair use, the retrieval of live pages to build chatbot answers may be treated differently, and that is the activity closest to how publishers’ current reporting reaches readers through AI tools.

The market dilution argument matters beyond newspapers. Brands and publishers that rely on content marketing and search traffic now compete with machine-generated pages as well, and they face the same question as the Times: does competition from machine-generated content count as harm the law recognizes? The briefs OpenAI selected argue that it does not. The Copyright Office’s 2025 report suggested it might.

The privacy dispute matters for a different reason. Conversations that users typed into a chatbot became discovery material in a copyright case, de-identified and produced under a protective order, and then supplied the regurgitation statistics both sides now cite. For an industry building advertising and measurement products around conversational AI, that is a concrete example of chat logs leaving the product and entering a courtroom.

None of the October 30 material changes the legal position. Summary judgment motions from both sides remain before Judge Stein. OpenAI’s page, and the 20 excerpts it now promotes, are its attempt to frame the question before the court answers it.

TimelineDecember 27, 2023 – The New York Times sues OpenAI and Microsoft for copyright infringement; OpenAI says it learned of the suit by reading the Times.February 2024 – OpenAI files its motion to dismiss, accusing the Times of manipulating prompts to produce regurgitation.November 7, 2024 – A judge dismisses Raw Story Media and AlterNet’s DMCA claims against OpenAI.November 22, 2024 – OpenAI writes to the court disputing the Times’ claim that data was destroyed during discovery.December 2, 2024 – Magistrate Judge Ona T. Wang orders OpenAI to produce employees’ work-related social media messages.April 4, 2025 – The court issues an opinion dismissing several of the Times’ claims.April 24, 2025 – Ziff Davis sues OpenAI in Delaware.May 2025 – The Times demands 1.4 billion ChatGPT conversations, according to OpenAI; the US Copyright Office releases its pre-publication AI training report.June 23, 2025 – Judge William Alsup rules training is fair use in Bartz v. Anthropic but rejects the pirated library defense.June 25, 2025 – Judge Vince Chhabria grants Meta summary judgment in Kadrey v. Meta.July 29, 2025 – OpenAI tells the court the Times deleted evidence of its internal use of OpenAI’s models.September 5, 2025 – Anthropic agrees to a $1.5 billion settlement in Bartz.November 7, 2025 – Wang orders OpenAI to produce 20 million de-identified ChatGPT logs.November 13, 2025 – Wang denies OpenAI’s stay request.November 24 and December 10, 2025 – OpenAI appeals the logs order to the district judge.December 15, 2025 – The court dismisses a number of Ziff Davis’s claims, including trademark and DMCA allegations.December 16, 2025 – OpenAI says it is complying with the logs order under de-identification and access controls.January 2026 – Stanford researchers report extracting copyrighted books from production AI models.August 5, 2026 – The Times reports Q2 2026 digital advertising revenue of $114.0 million and $4.6 million in AI litigation costs.August 6, 2026 – The court issues orders narrowing claims in the Times and Ziff Davis cases.August 10, 2026 – Meredith Kopit Levien sets out the Times’ AI licensing conditions on Bloomberg’s Odd Lots.September 1, 2026 – The Justice Department files a statement of interest backing fair use for AI training.September 4, 2026 – OpenAI, Microsoft and five news plaintiff groups move for summary judgment.September 23, 2026 – OpenAI moves to strike a study whose funding by plaintiffs’ counsel was not disclosed.October 6, 2026 – The court grants the motion to strike, and the plaintiffs withdraw the study, according to OpenAI.October 29, 2026 – The Wikimedia Foundation asks the court for leave to file a brief in support of neither party, backing fair use for training but not for ChatGPT’s retrieval of web pages.October 30, 2026 – OpenAI publishes “Fair use benefits everyone” with excerpts from 20 filings and updates its case page; it also publishes a post on AI training preferences for websites.Summary

Who: OpenAI, defendant in the copyright case brought by The New York Times, along with the 20 amici it quoted, among them the United States government, the Electronic Frontier Foundation, library and disability groups, law professors, computer scientists, Khosla Ventures and former lawmakers Rick Santorum and Rick Boucher. The Times, Ziff Davis and other news plaintiffs are on the other side.

What: OpenAI published “Fair use benefits everyone,” a post collecting excerpts from 20 filings that argue AI training is fair use, and rewrote its case page to call the Times’ lawsuit a pursuit of “an undeserved payday.” The page adds new 2026 entries on an order narrowing claims, summary judgment motions and a granted motion to strike a study that plaintiffs’ counsel had funded without disclosure.

When: The post was published on Friday, October 30, 2026. The litigation began on December 27, 2023, summary judgment motions were filed on September 4, 2026, and the motion to strike was granted on October 6, 2026, according to OpenAI.

Where: On OpenAI’s website, about multidistrict litigation 25-md-3143 before Judge Sidney H. Stein in the US District Court for the Southern District of New York.

Why: Judge Stein’s ruling on fair use will decide whether AI developers need licenses to train on news content, which sets the bargaining position of every publisher negotiating with AI companies and the exposure of AI developers to statutory damages calculated per article.