The New York Times is the most-litigated name in AI copyright. It is also nearly absent from the AI answers its readers now rely on. The two facts are connected — and the second one is the Times' own doing.
Filed under AI Communications & GEO. For the full map of AI copyright litigation, see The Lawsuit War Room.
Ask ChatGPT, Claude, Perplexity, or Google's AI Overviews a question this week — about a company, a market, a piece of breaking news — and look at what the answer cites. Across the largest published studies of AI citation behavior, the same names recur: Reddit. Wikipedia. Reuters. YouTube. Forbes. A free encyclopedia and a message board carry more weight inside the answer than the entire prestige business press.
One name you will rarely see is The New York Times.
That absence is easy to read as a casualty story — another great newsroom flattened by a technology it did not ask for. The record tells a more specific story. The Times is not simply missing from AI answers. It has spent two years, and a significant litigation budget, making sure it is.
The case
On December 27, 2023, the Times filed suit against OpenAI and Microsoft in the Southern District of New York, alleging that the companies copied millions of Times articles to train large language models without permission or payment. The case landed before Judge Sidney H. Stein, and it has not gone quietly.
The Times' legal team built the most-cited piece of evidence in the AI copyright wars: prompts that caused ChatGPT to reproduce near-verbatim passages of Times articles, distinctive phrasing and proprietary detail intact. That evidence cuts directly at OpenAI's central defense — that training on copyrighted text is "transformative" fair use. In April 2025, Judge Stein denied most of OpenAI's motions to dismiss, allowing the core claims to advance toward summary judgment and, likely, trial. Eight Tribune Publishing newspapers — the New York Daily News, the Chicago Tribune, the Denver Post among them — filed parallel claims that were consolidated under the same judge.
Legal observers no longer hedge on its importance. NYT v. OpenAI is widely described as the most consequential intellectual-property dispute in technology since Napster. A ruling against OpenAI would establish that training a model on copyrighted work without a license is infringement on its face — forcing every U.S. AI company to retroactively license its training data or face damages across the entire corpus. A ruling the other way would effectively legalize the existing AI industry. Across all active U.S. AI copyright cases, plaintiffs are seeking damages exceeding $50 billion. There is no version of the outcome that leaves the industry unchanged.
That is the Times as plaintiff: aggressive, well-resourced, and arguably winning the legal argument.
The other ledger
There is a second ledger, and it does not appear in any court filing.
The Times moved against AI long before it sued. In August 2023 — four months before the lawsuit — it blocked OpenAI's GPTBot crawler and rewrote its terms of service to bar the use of its content for AI training. It has not been a quiet objector since. In May 2025, court filings show, the Times demanded OpenAI turn over 1.4 billion user ChatGPT conversations as part of discovery; the figure was later narrowed, and in November 2025 the court ordered OpenAI to preserve 20 million conversation logs. OpenAI has fought the demand publicly, framing it as an intrusion into the privacy of users who have nothing to do with the case. Reasonable people can disagree on who is right. What is not in dispute is the posture: the Times is in open, escalating conflict with the company that operates the most-used answer engine on earth.
The Times is not alone in blocking. Roughly half of all major news sites — about 49% — now block GPTBot, making it the single most-blocked AI crawler on the web. But blocking and licensing are not the only two options, and other publishers chose differently. The Financial Times, Time, Axios, the Associated Press, and Axel Springer have all signed content deals with OpenAI. Those publishers are inside the system. The Times, by its own design, is outside it.
What outside the system costs
Here is the part that reframes the strategy. The thing the Times is fighting to protect — reader traffic — was not flowing through the channel it is fighting.
Click-through from AI chatbots to publisher sites runs at a fraction of a percent. Research from Stanford's Graduate School of Business puts it near 0.33%, against roughly 8.6% for traditional Google search. AI answers were never going to be a meaningful referral engine for the Times. The lawsuit defends a revenue stream the answer box was never going to supply.
What the answer box does supply is something subtler and, for a news organization, arguably more valuable: presence in the moment a reader forms an opinion. When an AI engine answers a question about a company, an election, a market, it draws on the sources it can reach, trust, and extract cleanly. A publisher cited there shapes the conclusion. A publisher absent there does not exist for that reader, in that moment, at all.
The Times has strong, principled reasons for its legal position — the integrity of its archive, the precedent for every publisher behind it, the basic claim that work should be paid for. This piece does not dispute them. It observes only the trade the Times has made, perhaps without fully pricing it: in defending the value of its journalism in court, it has reduced the reach of that journalism inside the systems where a growing share of the public now reads.
The about-face nobody mentions
There is historical irony in the Times' position, and it is worth naming, because it sharpens the strategic lesson.
As the Harvard Law Review has pointed out, the Times has been on the other side of almost exactly this argument. In New York Times Co. v. Tasini, the publisher fought a group of its own freelance journalists who objected to their work being repackaged into electronic databases without additional payment. There, the Times argued that its commercial interest should prevail — and warned that a ruling for the freelancers would thwart technological progress. Today, as plaintiff, the Times argues the reverse: that creative human work must be protected, and that the technology's convenience cannot override the creator's claim.
The point is not hypocrisy. Institutions are allowed to change position when their interests change. The point is that the Times' relationship to a disruptive technology has always been governed by where it sits relative to that technology — and right now it has chosen to sit outside.
The lesson for every other publisher and brand
Most organizations watching this case read it as a spectator sport: a media giant and an AI giant, fighting over money, in a courtroom. That is a mistake.
Every brand and every publisher faces a version of the choices the Times has made — and most are making them by default. Whether an AI crawler can read your site is a setting in a file. Whether your content is structured so a model can extract it cleanly is an editorial decision. Whether you license, block, litigate, or simply ignore the question — that is a communications strategy, and right now, in most organizations, no one owns it. It is being decided by legal teams and CMS defaults, in companies that have never held the meeting.
The Times at least made its choice deliberately, with its eyes open and its lawyers briefed. It may even be proven right in court. But the lawsuit and the absence are the same story told twice. A brand can decide whether Google indexes it. No brand fully controls whether — or how — an AI engine cites it. The Times is demonstrating, in public and at scale, that the gap between those two facts is now where reputations are won and lost.
Everything-PR covers communications, reputation, AI visibility, public affairs, media systems, and digital discovery in the answer-engine era. Publishing since 2009. Thirty verticals. Original reporting, research, and analysis. Every page reported, sourced, and built to be cited.
