DataPacts

News Publishers v. Perplexity (New York Times, Chicago Tribune, CNN): Claims and Status

Checked

At a glance

The New York Times, the Chicago Tribune and CNN are suing Perplexity AI, Inc. in three coordinated copyright and trademark cases in the U.S. District Court for the Southern District of New York before Judge Loretta A. Preska (Nos. 1:25-cv-10106-LAP, 1:25-cv-10094-LAP and 1:26-cv-04427-LAP), alleging that Perplexity's answer engine copies their journalism into a retrieval index and reproduces it in answers.

Independent reference. Every buyer is covered on the same terms, every figure carries its source and check date, and no buyer pays to be listed or placed higher.

Request an independent introduction

Tell us what data your company holds. We reply with the buyers that fit and what their terms leave out. Your details reach a buyer only after you agree to a named introduction.

IndependentYour details stay with us until you agreeNot selling is a valid outcome

The Times complaint is dated December 5, 2025, the Tribune complaint was filed on December 4, 2025 and the CNN complaint on May 28, 2026. Each pleads the same five counts, and Perplexity's partial motions to dismiss leave the input-copying count untouched.

As of November 17, 2026, no court has ruled on the merits or on Perplexity's motions, and every claim is an allegation. This page belongs to the independent reference on AI training data licensing, and DataPacts is not affiliated with any party and gives no legal advice.

Every date and docket entry comes from the federal dockets and the Times' quarterly SEC filing, and how figures are sourced and dated is set out on the methodology page.

The Three Publisher Suits at a Glance: Courts, Dockets, Filing Dates and Status

All three suits are pending, with Perplexity's motions to dismiss part of each case fully briefed and undecided as of November 17, 2026. The table compares the three cases as of 2026-11-17.

Case Court and judge Docket Filed Plaintiffs Counts Status Source
The New York Times Co. v. Perplexity AI, Inc. S.D.N.Y., Senior U.S. District Judge Loretta A. Preska 1:25-cv-10106-LAP 2025-12-05 The New York Times Company; Wirecutter, Inc. and The Athletic Media Company added 2026-03-20 5 (copyright x3, trademark x2) partial motion to dismiss (Counts II to V) fully briefed 2026-06-09; discovery under way Dkt. 1, 3, 44, 68; NYT 10-Q
Chicago Tribune Company, LLC v. Perplexity AI, Inc. S.D.N.Y., Judge Preska 1:25-cv-10094-LAP 2025-12-04 Chicago Tribune Company, LLC 5 partial motion to dismiss (Counts II, III) fully briefed 2026-06-09; source code protocol 2026-10-21 Dkt. 1, 37, 52, 66
Cable News Network, Inc. v. Perplexity AI, Inc. S.D.N.Y., Judge Preska 1:26-cv-04427-LAP 2026-05-28 Cable News Network, Inc. 5 motion to dismiss (output, secondary liability, trademark) fully briefed 2026-10-23, oral argument requested Dkt. 1, 31, 42

Defendant in all three: Perplexity AI, Inc., a Delaware corporation based in San Francisco. No ruling on the merits as of 2026-11-17.

The Five Counts Each Publisher Pleads Against Perplexity

Each complaint pleads the same five counts: copyright infringement through copies made as inputs for a retrieval index, copyright infringement through answers, contributory and vicarious infringement, false designation of origin and dilution, and trademark infringement. The table sets the counts beside the plaintiffs and Perplexity's motions, as of the amended complaints of 2026-03-20.

Count Statute What it targets (publishers' allegation) NYT Tribune CNN Challenged by Perplexity's motion Source
I: copyright, inputs 17 U.S.C. 106(1) copies into retrieval databases or indexes yes yes yes no complaints, Count I
II: copyright, outputs 17 U.S.C. 106(2) answers with verbatim, near-verbatim or derivative text yes yes yes yes Count II; Times Dkt. 58; CNN Dkt. 31
III: contributory and vicarious infringement 17 U.S.C. 106, 501 users' and licensees' infringement, pleaded in the alternative yes yes yes yes Count III
IV: false designation of origin and dilution 15 U.S.C. 1125(a), (c) answers attributing invented or incomplete text to the marks yes yes yes Times and CNN: yes; Tribune: no Count IV
V: trademark infringement 15 U.S.C. 1114 hallucinated articles under the marks yes yes yes Times and CNN: yes; Tribune: no Count V

The same counsel filed all three complaints, and the Times and Tribune motions are briefed in consolidated papers. The Times' first amended complaint of March 20, 2026 added Wirecutter and The Athletic as plaintiffs.

The Tribune and CNN plead the trademark counts too, so none of the three suits is a copyright-only case, and all statements here are allegations.

The publishers allege that Perplexity copies their articles, directly or through third parties' crawlers, into databases or indexes used for retrieval-augmented generation (Count I, 17 U.S.C. 106(1)), then reproduces or summarizes them in answers (Count II, 17 U.S.C. 106(2)). Count III applies in the alternative if users, not Perplexity, are the direct infringers.

Scale is pleaded as allegation: the Times cites over 10 million registered works, the Tribune registered copyrights for "millions" of articles, and CNN says Perplexity copied "over 17,000" CNN stories, videos, images and other works (Exhibit A lists them).

The trademark counts rest on answers that attribute invented or incomplete text to the mastheads. CNN adds that Perplexity's chatbot answered the question "What is comet plus?" by placing CNN in a "premium news bundle".

Before the Suits: Letters, robots.txt Blocks and CNN's Abandoned Term Sheet

Each publisher pleads that it told Perplexity to stop before suing. The Times alleges letters of March 29, 2024 (demanding a stop to all use, "including but not limited to use in training, unless and until we reach a negotiated agreement that includes fair compensation"), May 10, 2024, October 2, 2024 and July 23, 2025, and that it disallowed PerplexityBot in robots.txt on February 27, 2024 and hard-blocked it on November 21, 2024.

It says Perplexity-User was blocked both ways on July 22, 2025 and that it recorded over 175,000 access attempts in August 2025. Perplexity's own statements, checked 2026-11-09, say PerplexityBot is "not used to crawl content for AI foundation models" and Perplexity-User "generally ignores robots.txt", as covered on the page for PerplexityBot and Perplexity-User.

The Tribune alleges an October 14, 2025 letter from its parent's counsel, and a reply of November 10, 2025 in which Perplexity's counsel wrote that Perplexity "does not train any models" with that content and, where robots.txt is in place, has not accessed protected content.

The Tribune calls the reply incorrect and alleges over 16,700 access attempts in six months.

CNN alleges a term sheet effective October 1, 2025 for a Comet Plus partnership, with a contract to follow within 60 days and paywalled access for compensation.

It says the term sheet was terminated effective November 24, 2025 after disagreement that included limits on CNN content in answers, and that its letter of December 10, 2025 went unanswered. The terms are not stated here.

What the Publishers Ask the Court to Order

The publishers seek statutory and actual damages, restitution of profits, a permanent injunction, and costs and fees, and each demands a jury trial. CNN adds treble damages. No complaint states an amount, and nothing has been awarded.

Perplexity's Partial Motions to Dismiss: Outputs, Secondary Liability and Trademarks

Perplexity has not moved to dismiss the input-copying count in any of the three cases; its motions target the answer-based copyright count, the secondary-liability count and the trademark counts.

In the Times and Tribune cases it filed a first consolidated motion on February 27, 2026, the plaintiffs amended on March 20, 2026, and the renewed motion of April 17, 2026 was opposed on May 15 and answered by a reply of June 9, 2026.

In the CNN case the memorandum is dated August 21, 2026, CNN opposed on September 25 and Perplexity replied on October 23, 2026.

Perplexity argues that answers are generated by users' prompts, not by volitional conduct of its own (citing Cartoon Network LP v. CSC Holdings, 536 F.3d 121 (2d Cir. 2008)), that contributory liability fails under Cox Communications v.

Sony Music Entertainment (2026), that it has no direct financial interest in any infringement, and that its references to the marks are nominative fair use of marks that, for Wirecutter and The Athletic, are not famous. The plaintiffs answer that Perplexity selects and supplies the text.

Perplexity calls fair use "the core issue" for the claims that remain. Both motions are fully briefed with oral argument requested, and no ruling is located as of November 17, 2026. The defense Perplexity reserves is explained on the page on fair use and AI training in US law.

News Publishers v. Perplexity Docket Entries, 2025 to 2026

Fifteen docket entries across the three cases run from the Tribune complaint of December 4, 2025 to the plaintiffs' discovery reply of November 10, 2026. The table lists them as of 2026-11-17.

Date Case and entry What it is or decided Source
2025-12-04 Tribune Dkt. 1 complaint, five counts, jury demand RECAP archive PDF
2025-12-05 Times Dkt. 1, Dkt. 3 complaint dated 2025-12-05; related-case statement names Dow Jones (24-cv-07984) as the earlier case, and refers to Britannica and the Tribune RECAP archive PDFs
2026-02-27 Times Dkt. 40, 43; Tribune Dkt. 34 Perplexity's first consolidated motion to dismiss RECAP archive PDFs
2026-03-20 Times Dkt. 44; Tribune Dkt. 37 first amended complaints; Wirecutter and The Athletic join the Times RECAP archive PDFs
2026-04-17 Times Dkt. 57, 58; Tribune Dkt. 42 renewed consolidated partial motion to dismiss RECAP archive PDFs
2026-05-15 Times Dkt. 59; Tribune Dkt. 43 plaintiffs' consolidated opposition RECAP archive PDFs
2026-05-28 CNN Dkt. 1 complaint, five counts, Exhibit A works list RECAP archive PDF
2026-06-09 Times Dkt. 68; Tribune Dkt. 52 Perplexity's reply; oral argument requested RECAP archive PDFs
2026-08-07 Times Dkt. 74 Perplexity letter on retrieval snapshots of November 2024 and May 2025 RECAP archive PDF
2026-08-21 CNN Dkt. 31 (motion ECF 30) Perplexity's motion to dismiss CNN's output, secondary-liability and trademark claims RECAP archive PDF
2026-08-31 Tribune Dkt. 60 stipulated protective order with an Inspection Data tier RECAP archive PDF
2026-09-25 CNN Dkt. 37 CNN's opposition RECAP archive PDF
2026-10-21 Tribune Dkt. 66; Times Dkt. 81 (proposed form) source code protocol RECAP archive PDFs
2026-10-23 CNN Dkt. 42 reply filed, briefing complete; oral argument requested RECAP archive PDF
2026-11-03 to 2026-11-10 Tribune Dkt. 71 and 75, Times Dkt. 89; Times Dkt. 93; Times Dkt. 94 plaintiffs seek an order on user-log costs and training discovery; Perplexity opposes (2026-11-06); plaintiffs reply (2026-11-10) RECAP archive PDFs

Entries come from the RECAP archive of the three dockets, read 2026-11-17. The archive is incomplete, and no later entry or ruling was located.

Discovery Before Judge Preska: Protective Order, Source Code and RAG Snapshots

The Times and Tribune cases share discovery orders: a stipulated protective order entered on August 31, 2026 with a separate inspection-data tier, and a source code protocol entered on the Tribune docket on October 21, 2026 (Dkt. 66).

Perplexity's letter of August 7, 2026 states that its two archival copies of retrieval data are snapshots of all such data on a date in November 2024 and in May 2025, and that click-through data is kept and preserved separately.

The plaintiffs' letter of November 3, 2026 disputes who bears the cost of inspecting user activity logs, and quotes Perplexity's estimate of at least USD 12,000 a month for the May 2024 to May 2025 snapshot.

Perplexity answers on November 6, 2026 that it seeks no hosting costs for existing snapshots, only for a new one. No court ruled on the dispute in the record read, and the CNN docket shows no discovery order.

Is Training in the Case? The November 2026 Discovery Dispute

Perplexity's position, in its letter of November 6, 2026, is that "This is not a 'training' case": the complaints plead copying into a retrieval index and answers built from it, not training, post-training or fine-tuning on the publishers' content.

It therefore refuses the plaintiffs' requests for production on those topics (Dkt. 93-1) and adds that the output claim is subject to its pending motion to dismiss.

The plaintiffs reply on November 10, 2026 that their amended complaints allege Perplexity fine-tuned Sonar models to "optimize for RAG", including to browse news, and that "if it does not train on Plaintiffs' content, it should say so".

Their reply sits beside the Tribune's allegation of November 10, 2025 that Perplexity's counsel wrote that Perplexity "does not train any models" with its content, a statement the Tribune disputes. The court has not ruled, and no filing states that Perplexity did or did not train on this content.

The case tracker keeps these suits outside its training table until a court record shows training. The earliest suit's retrieval reasoning is on the page for Dow Jones and NY Post v. Perplexity, where the complaint likewise targets retrieval copies.

What the Suits Show Publishers Weighing a License With an Answer Engine

The CNN complaint records a license that nearly happened: a term sheet for paid access that ended over how much of CNN's content Perplexity's answers could use.

All three complaints cite Perplexity's Publishers' Program and its licensing deals as evidence of a market for licensing content for AI. The terms of that program and the licenses on record are on the profile of Perplexity as a data buyer, and this page states none of them.

CNN names one of the "multiple issues" in dispute, limits on Perplexity's use of CNN content in answers, and lists no others. The record shows no outcome, damages or price for the talks.

Publisher license structures, from training licenses to display and answer licenses, are compared on the page on news publishers licensing content to AI companies.

The three publisher suits are the most recent court cases against Perplexity in the tracker, all about copying for AI answers or circumventing access controls rather than a pleaded training claim.

Every tracked case with docket and status is on the case tracker of AI copyright lawsuits over training data, and the table lists the Perplexity cases and the Times' suit against OpenAI as of 2026-11-17, with the status date the page states.

Case Court and docket Claim type Status Source
Dow Jones and NY Post v. Perplexity S.D.N.Y. 1:24-cv-07984-KPF, filed 2024-10-21 copyright and trademark, answer engine discovery; motion to compel granted in full 2026-10-29 (Doc. 227), as of 2026-11-09 Doc. 1, 227
Britannica and Merriam-Webster v. Perplexity S.D.N.Y. 1:25-cv-07546-JLR-SLC, filed 2025-09-10 copyright and trademark, answer engine discovery; Opinion and Order 2026-10-27 (Doc. 152), as of 2026-11-09 Doc. 152
Reddit v. Perplexity, SerpApi and others S.D.N.Y. 1:25-cv-08736-PAE, filed 2025-10-22 circumvention of access controls (DMCA 1201) and related claims pending; motions to dismiss of SerpApi and Perplexity predominantly denied 2026-07-31 (Dkt. 104), as of 2026-11-17 Dkt. 1, 104
The New York Times v. Microsoft and OpenAI S.D.N.Y. 1:23-cv-11195-SHS-OTW, filed 2023-12-27, MDL No. 3143 copyright, training summary-judgment stage, no ruling located, as of 2026-11-09 MDL No. 3143 transfer order

The Times' suit against Microsoft and OpenAI is a training case in MDL No. 3143 and is not coordinated with the Perplexity suit.

More About the Publisher Suits Against Perplexity

Decided Versus Still Open in the Publisher Suits

Only procedure is decided so far: protective and source code orders, nothing on liability. The lists separate the two as of November 17, 2026.

  • Decided: the cases sit before Judge Preska, with joint Times and Tribune briefing.
  • Decided: the protective order was entered on 2026-08-31.
  • Decided: the source code protocol was filed on 2026-10-21.
  • Decided: both motions to dismiss are fully briefed.
  • Open: the rulings on Perplexity's motions.
  • Open: whether training discovery is allowed.
  • Open: infringement and fair use.
  • Open: a trial date, which no filing sets.

Frequently Asked Questions About the Publisher Suits Against Perplexity

Is the New York Times suing Perplexity over AI training?

The Times' five counts target copies in a retrieval index and answers built from them, not training. Whether training is discoverable is disputed and undecided as of 2026-11-17.

Has the court ruled on Perplexity's motions to dismiss?

No ruling appears on the record as of 2026-11-17. The Times and Tribune motion was briefed by 2026-06-09 and the CNN motion by 2026-10-23.

Is this the same case as The New York Times v. OpenAI?

No. That suit is S.D.N.Y. 1:23-cv-11195-SHS-OTW, filed 2023-12-27, now part of MDL No. 3143, and it is about training.

Why did CNN's licensing talks with Perplexity end?

CNN alleges the term sheet of 2025-10-01 was terminated effective 2025-11-24 after disagreement that included limits on CNN content in answers.

When did Dow Jones sue Perplexity?

Dow Jones and the New York Post sued on 2024-10-21 (1:24-cv-07984-KPF, Doc. 1). The "January 2025" in the Times complaint is the date of Dow Jones' second amended complaint, filed 2025-01-28.

This page is information, not legal advice.