You have probably spent a year making your pages easy for AI to quote. An adviser to Europe's top court has now described where quoting turns into copying.

On November 3, 2026, Advocate General Maciej Szpunar told the EU Court of Justice that a chatbot which retrieves a press article and reuses it in an answer may infringe the press publisher's right "in certain circumstances." Two things decide it: how the chatbot got the article, and how closely the answer follows it. Training, he said, is as a rule permitted under the text and data mining exception. The case is Like Company v Google Ireland, about Gemini.

My read: the test sits close to the self-contained passages AEO guides tell you to write. The opinion never mentions AEO, and that link is my inference.

The court laid the opinion out in a press release the same day.

Like Company, a Hungarian operator of web portals, sued Google in Budapest. It says Google used its articles to train Gemini. It also says Google reproduced them and made them available through the chatbot between June 13, 2023 and May 31, 2024. The Hungarian court sent four questions to Luxembourg. Two cover training and two cover the answers.

Szpunar advised the court to declare the training questions inadmissible. He found no evidence that Like Company's articles were used to train Gemini, and he said training is not a form of "online use" of press publications. For completeness, he added that training is "as a rule, freely permitted under the text and data mining exception." Reuters reports that Google welcomed that part, saying the opinion recognises that "EU law permits AI training on publicly available information."

The test for answers

On the answers, the Court of Justice press release summarising AG Szpunar says: "when an LLM-driven chatbot retrieves a publication from a press publisher's website, reuses that publication to generate an answer, and provides that answer to the user, the press publisher's right may be infringed in certain circumstances".

The right has limits. It protects the specific form of expression of a press publication and "does not protect the mere facts reported in press publications." Its aim, the release says, is to protect the investment behind a publication, "not to give publishers a monopoly over the news" itself.

Within those limits, two questions decide the case. The first is access. Did the chatbot get around a paywall or a similar technical measure meant to restrict access or scraping? The second is output. Does the answer only report facts from the article, or does it reuse the article's form of expression in a recognisable way? The release puts the line at the point where the reader is "entirely dispensed from accessing the original publication."

The right of making available to the public is, as a rule, infringed where that form of expression is systematically recognisable in the chatbot's answers to related queries. The provider is liable where such uses are inherent to its business model. It is also liable where it "failed to put in place appropriate measures in order to counter those acts credibly and effectively."

Two readings, and a court still to decide

Google and the publishers read the same opinion differently. A Google spokesperson told Reuters that "Gemini learns linguistic patterns to create original outputs rather than replicating content, supported by extensive publisher controls that give rightsholders clear choice and the ability to opt out." Angela Mills Wade, executive director of the European Publishers Council, said that "Publishers should not have to choose between search visibility and protecting their rights."

Early reaction on X was tentative. Legal commentator Technollama wrote, before reading the full text, that it "seems like a good opinion for the most part." SEO consultant Glenn Gabe called it a "gray area," which is a first reaction and not legal analysis.

Nothing is settled. The judges are now beginning their deliberations, and judgment will come later. Reuters says the court follows the majority of its advisers' non-binding opinions and will rule in the coming months.

What I take from it

The right covers press publications only. A SaaS blog or a brand site gains nothing from it. If you do publish news in the EU, the live question is what an answer does with your wording at retrieval time.

My inference is that short, self-contained passages are the format most likely to be reproduced in a recognisable way, because they are written to be lifted whole. A publisher who wants citations and also wants protection for its wording is asking for two things that pull against each other. I think that tension is the real story here, and nobody has resolved it. None of this is legal advice.

Free newsletter

The Sherpa Brief

One email a week on what changed in AI search, what it means, and what to do about it. Written for marketers, SEOs, and publishers.

Free. Unsubscribe in one click.