🔍 Read the full analysis: AI Copyright In Europe: A Court Adviser’s View On Publishers’ Rights on ThorstenMeyerAI.com
Get business pricing on tech for your team
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
On 3 November 2026, Advocate General Maciej Szpunar advised the EU Court of Justice that chatbots retrieving and reusing news articles may infringe publishers’ rights in specific circumstances. He said the test concerns unlawful access or answers that reproduce an article’s recognisable expression—not merely the reuse of its facts. His opinion is non-binding, and he said the conditions did not appear to be met in the dispute over a Gemini summary of a Hungarian article.
Advocate General Maciej Szpunar advised the European Union’s Court of Justice on 3 November that an AI chatbot may infringe press publishers’ rights if it accesses an article unlawfully or reproduces its recognisable expression so extensively that users can forgo the original. The opinion concerns Google’s Gemini chatbot and a Hungarian publisher’s lawsuit, but it is not a ruling, and Szpunar said the test did not appear to be satisfied on the facts presented.
The case, C‑250/25, Like Company v Google Ireland, began after Like Company, which operates Hungarian news websites, sued Google in a Budapest court. The publisher alleged that Google used its articles to train Gemini and repeatedly reproduced and made them available through chatbot answers between 13 June 2023 and 31 May 2024. Those are the publisher’s claims, not findings established by the EU court.
The dispute includes a summary of an article published in July 2023 about a singer’s reported plan to bring dolphins to Lake Balaton. Gemini produced a Hungarian summary after a user asked about the article and supplied the site’s domain. Szpunar said chatbot answers may cross the line if the service bypasses a barrier such as a paywall or anti-scraping measure, or if its response preserves the article’s distinctive expression to the point that a reader need not visit the publication.
By contrast, repeating the facts alone is not enough under the adviser’s proposed test. He said the example did not appear to involve access barriers or recognisable wording, and Google appeared to have safeguards. The Hungarian court must assess the evidence and apply the EU court’s eventual interpretation.
Infographic 01 · Where the line runs
When a chatbot answer infringes a publisher's right
Advocate General Szpunar's test in Case C‑250/25 for chatbots that retrieve news articles in real time and reuse them in their answers.
Stays free
- Reporting the same facts. Nobody owns the news itself (pts 199–200).
- A short factual summary, presented differently (pt 208).
- An overview built from several sources, restructured, with links back (pts 208, 234).
- Individual words or very short extracts (pt 201).
Can infringe
- Getting around a paywall or anti-scraping block to fetch the article (pt 188).
- An answer that keeps the article’s structure and phrasing, even paraphrased, so readers can skip the original (pts 202, 205, 207).
- Systematically recognisable articles in answers to related questions: making available to the public (pts 213, 240).
The test, step by step
For a chatbot that fetches a news article live (RAG) and answers a user
- Is it an online use by a service in the EU?Yes for RAG answers to EU users, wherever the servers are (pts 165–169).
- Did the bot get in lawfully?No circumvention of effective technical measures (pt 188).
- Is the article’s form of expression recognisable in the answer, to the point of substitution?Side-by-side comparison; facts do not count (pts 196–209).
- Is the provider responsible?If inherent to its business model, or no reasonable measures such as respecting blocks and output filters (pts 221–222).
- Temporary copies during retrieval are then exempt under Art. 5(1) InfoSoc; the TDM exception does not apply to RAG (pts 179–189).
An Advocate General’s opinion, not a judgment. General information, not legal advice.
Sources: Opinion of AG Szpunar, C‑250/25 (provisional text), CURIA press release No 155/26 (3 Nov 2026). Point numbers refer to the provisional text.
Powered by Thorsten Meyer AI
A Narrower Test for Chatbot Reuse
The opinion could help define how the EU’s press publishers’ right applies to AI systems that search the web and generate answers, rather than simply displaying links or short extracts. If the Court adopts Szpunar’s approach, publishers may have grounds to challenge chatbot use that gets around access controls or substitutes for the original by reproducing its distinctive wording.
That would not give publishers ownership over news facts or a general veto over AI summaries. For publishers, the distinction could shape negotiations over licences and the technical controls used to restrict access. For AI companies, it points to risks around how their systems retrieve material and how closely generated answers track source language. For readers, the eventual rules may affect whether chatbots provide summaries, links, or less detail from protected publications.
These are possible consequences, not settled outcomes. An Advocate General’s opinion is independent legal advice to the Court; the judges may follow it, modify its reasoning or reach a different conclusion.
Products Worth Considering
Logitech MX Brio Ultra HD 4K Webcam, Graphite
Samsung Odyssey G4 25-inch FHD Gaming Monitor
Bose SoundLink Plus Portable Bluetooth Speaker, Wireless Outdoor Speaker, Up to 20 Hours Battery Life, Waterproof and Dustproof, Black
As an affiliate, we earn on qualifying purchases.
How the Hungarian Case Reached Luxembourg
EU law has recognised a related right for press publishers since 2019. Article 15(1) of the Copyright in the Digital Single Market Directive gives publishers rights over certain online uses of their publications. Szpunar’s opinion notes that lawmakers had services such as Google News in mind. The case asks whether that framework also covers chatbot answers built from articles retrieved in response to a user’s prompt.
The Budapest regional court sent four questions to Luxembourg: two concerning AI training and two concerning chatbot responses. Nine EU governments and the European Commission submitted written observations, and the hearing took place on 10 March 2026. The EU Court of Justice described the dispute as the first generative AI case to reach it.
On model training, Szpunar’s principal conclusion was procedural: he advised the Court to treat the two questions as inadmissible because the record did not establish that Like Company’s articles were used to train Gemini. He also considered training outside the publishers’ right at issue because that right concerns online use by services. Separately, and without making it necessary to resolve the dispute, he said text-and-data-mining rules generally cover training in principle, subject to rights holders’ ability to opt out in a machine-readable form. That discussion was not a finding that Google trained Gemini on the publisher’s articles.
““the first generative AI case to reach the Court of Justice””
— The Court of Justice of the European Union, press release No 155/26
The Court Has Not Ruled Yet
The opinion is not binding, and the Court of Justice has not announced when it will deliver its judgment. It remains unknown whether the judges will adopt Szpunar’s proposed test, including his distinction between reporting facts and reproducing recognisable expression.
The factual record also leaves key issues for the national proceedings. The publisher alleged that its articles were used in training, but Szpunar said the material before the Court did not support that claim. The Hungarian court will need to determine the relevant facts, including how the article was accessed and whether the answer reproduced protected expression. The opinion’s general discussion of text-and-data mining does not resolve those questions or establish that the publisher’s work was included in Gemini’s training data.
A Judgment Before the Case Returns
The Court of Justice will issue its judgment in response to the Hungarian court’s questions; no date has been set. The judgment will provide the EU-law interpretation, after which the Budapest court will apply that interpretation to the evidence in Like Company’s case.
The timing is uncertain. The Court’s 2025 statistics put the average duration of preliminary-reference proceedings at 16.9 months, but that is a past average, not a forecast for this case. Until the judgment and national-court assessment, the adviser’s proposed limits remain guidance rather than a final determination of Google’s liability.
Key Questions
Has the EU Court of Justice ruled that Gemini infringed copyright?
No. Szpunar issued a non-binding opinion, not a judgment. He said the infringement conditions did not appear to be met in the example, while leaving the final decision to the Court of Justice and the Hungarian court.
When could a chatbot answer infringe publishers’ rights under the opinion?
Szpunar proposed two relevant circumstances: the chatbot accesses an article unlawfully, such as by bypassing a paywall or anti-scraping barrier, or it reproduces recognisable expression so extensively that users can forgo the original. Repeating the article’s facts alone would not meet that test.
Did the opinion establish that Google trained Gemini on Like Company’s articles?
No. The publisher alleged training use, but Szpunar advised treating those questions as inadmissible because the record did not establish that the articles were used for training. His separate discussion of training and text-and-data-mining rules did not prove use of these articles.
What happens after the Advocate General’s opinion?
The Court of Justice will issue a judgment interpreting EU law. The Hungarian court will then decide the national dispute using that interpretation and the evidence. No judgment date has been announced.
Source: ThorstenMeyerAI.com
