Across the book, academic, education and press sectors, publishers and media companies have turned to the courts to defend their copyrights against technology companies. The wave of litigation spans several distinct legal theories: direct copyright infringement through the copying of works to build training datasets; reproduction of protected content in AI outputs; and — in the antitrust context — the use of AI-generated summaries to divert traffic and revenue away from publishers.

Core arguments common to most cases

1. Direct copyright infringement. The mass downloading, pre-processing and ingestion of texts into training databases constitutes unauthorised reproduction.

2. Market displacement by AI products. AI models generate summaries or substitute texts that make purchasing the original work unnecessary, depriving publishers of direct revenue.

3. Destruction of the licensing market. Publishers emphasise that functioning, commercial licensing markets for AI training already exist; technology companies that obtain data illegally deliberately bypass this market.

The most significant proceedings are set out below.

Publishers & Scott Turow v. Meta Platforms (May 2026)

Pending

On 5 May 2026, five major publishing houses — Elsevier, Cengage Learning, Hachette Book Group, Macmillan and McGraw Hill — together with bestselling author Scott Turow (and his company S.C.R.I.B.E.) filed a putative class action against Meta and its founder and CEO Mark Zuckerberg in the U.S. District Court for the Southern District of New York (captioned Elsevier Inc. et al. v. Meta Platforms, Inc., Case No. 1:26-cv-03689). The plaintiffs allege that Meta systematically copied millions of copyrighted textbooks, scholarly articles and literary works — torrented from pirate sites such as LibGen and Anna's Archive — to train its Llama large language models, without licence or authorisation. According to Publishers Weekly, this is the first copyright infringement suit brought against an AI company by publishers as such.

Penguin Random House Verlagsgruppe v. OpenAI (March 2026)

Pending

On 27 March 2026, the German Penguin Random House Verlagsgruppe filed suit before the Munich Regional Court (Landgericht München I) against OpenAI Ireland Ltd., the European subsidiary that provides ChatGPT. The case concerns the children's book series Der kleine Drache Kokosnuss ("Coconut the Little Dragon") by author and illustrator Ingo Siegner, published under the group's cbj imprint. The publisher alleges that, even in response to simple prompts, ChatGPT reproduces Siegner's protected texts and illustrations in recognisable form — an indication that the works were used without authorisation to train the model and are now stored as so-called "memorization." The complaint further alleges that ChatGPT proactively offered to assemble a print-ready knockoff, including infringing cover art and blurbs, and gave instructions for uploading it to self-publishing platforms. The suit builds on a November 2025 ruling by the same Munich court, which found that ChatGPT had infringed copyright by reproducing protected song lyrics in a case brought by the German collecting society GEMA.

Publishers v. Anna's Archive (March–May 2026)

Concluded — default judgment

On 6 March 2026, thirteen trade, educational and professional/scientific publishers — led by Apress Media and backed by the Association of American Publishers (AAP) — sued the shadow library Anna's Archive in the Southern District of New York (captioned Apress Media, LLC et al. v. Anna's Archive, Case No. 1:26-cv-01850, before Judge Jed S. Rakoff). Anna's Archive is a piracy platform, not an AI company, but the publishers alleged that it functions as an illegal supplier of high-quality book and journal data to AI developers in exchange for crypto payments. After the operators failed to appear, Judge Rakoff entered a default judgment on 19 May 2026, awarding the maximum statutory damages of USD 150,000 for each of the 130 works in suit (USD 19.5 million in total) and issuing a permanent injunction directing domain registries, registrars and hosting providers to cut off the site.

Author class actions with publisher involvement — Bartz v. Anthropic

Settled — in distribution

In Bartz v. Anthropic (N.D. Cal., Case No. 3:24-cv-05417), filed in August 2024 by authors Andrea Bartz, Charles Graeber and Kirk Wallace Johnson over the training of the Claude models, the court's class certification in July 2025 swept in not only authors but also publishers, estates and other holders of reproduction rights. In June 2025, Judge William Alsup held that training on lawfully acquired books was fair use, but that acquiring pirated copies from shadow libraries was not. The parties then agreed a landmark USD 1.5 billion settlement (roughly USD 3,000 per work, covering some 500,000 works); the final approval hearing took place on 14 May 2026. Because rights are often split, the settlement and distribution mechanics involve both authors and their publishers. Note that this matter is largely resolved rather than actively litigated.

The New York Times and other news publishers v. OpenAI & Microsoft

Pending

The foundational media case began in late December 2023, when The New York Times sued OpenAI and its partner Microsoft in the Southern District of New York, alleging that millions of its articles were copied without permission or compensation to build ChatGPT, and that the system can reproduce Times articles in near-verbatim form. A judicial panel subsequently consolidated this action with related suits — including those brought by the New York Daily News and other Alden/MediaNews Group papers (among them the Chicago Tribune), the Center for Investigative Reporting, The Intercept, and digital publisher Ziff Davis (owner of CNET) — into a single coordinated proceeding. Microsoft is a co-defendant in the Times action.

Most recently, on 9 July 2026, a group of 17 of these media organisations filed a motion seeking sanctions against OpenAI, alleging that the company acted in bad faith during discovery — falsely claiming for more than two years that it could not search its training datasets and output logs for copyrighted material, and withholding or destroying evidence central to the case. OpenAI denies the allegations. (The Times has reported spending more than USD 28 million on AI-related litigation.)

A separate strand of litigation targets Google — not primarily on copyright grounds, but on antitrust theories, arguing that Google's "AI Overviews" use publishers' content to generate on-page answers that cannibalise click-through traffic and the associated revenue.

Chegg, Inc. v. Google (February 2025)

Pending

Educational technology company Chegg sued Google and Alphabet in the U.S. District Court for the District of Columbia (Case No. 1:25-cv-00543), alleging that Google leverages its search monopoly to extract publisher content for AI Overviews without compensation and engages in unlawful "reciprocal dealing" — forcing publishers to supply content in order to appear in search results at all.

Penske Media Corporation v. Google (September 2025)

Pending

Penske Media — publisher of Rolling Stone, Variety, Billboard, The Hollywood Reporter, Deadline and more than twenty other brands — brought a similar antitrust action in the District of Columbia (Case No. 1:25-cv-03192), the first by a major US publisher over AI Overviews. Penske alleges click-through declines of up to 58% and argues that publishers are coerced into a choice between forfeiting search traffic entirely or watching it be cannibalised by AI summaries. Google moved to dismiss in January 2026, characterising AI Overviews as a lawful product improvement.

European Commission complaint (June 2025)

In parallel, a coalition including the Independent Publishers Alliance, the Movement for an Open Web and Foxglove Legal filed a formal antitrust complaint with the European Commission on 30 June 2025, accusing Google of abusing its dominance in search through AI Overviews and of breaching EU copyright rules, including publishers' neighbouring right under the DSM Copyright Directive.