🔍 Read the full analysis: How An EU Court Adviser’s Opinion Puts AI Chatbots In The Copyright Spotlight on ThorstenMeyerAI.com
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Advocate General Maciej Szpunar advised the EU Court of Justice that a chatbot may infringe press publishers’ rights if it accesses articles unlawfully or reproduces their distinctive expression closely enough to replace the original. He said those conditions do not appear to apply to the Gemini summary at issue, and the opinion is not binding. The court has not set a date for its judgment.
Advocate General Maciej Szpunar advised the European Union’s Court of Justice on November 3 that AI chatbots can, in some circumstances, infringe press publishers’ rights when they retrieve online articles and reuse them in answers. In Case C‑250/25, Like Company v Google Ireland, he said the key questions are whether a service accessed an article unlawfully or reproduced its expression so recognisably that users could avoid reading the original. His opinion is not a binding ruling, and he said the test does not appear to be met by the Gemini answer involved in this case.
The dispute began after Google’s Gemini chatbot summarised a Hungarian article about a proposal to bring dolphins to a pool on the Tihany peninsula at Lake Balaton. Like Company, which operates Hungarian news websites, sued Google in a Budapest court, alleging that it used the company’s articles to train Gemini and repeatedly reproduced or made them available through the chatbot from June 13, 2023, to May 31, 2024. Those are allegations described in the Court of Justice’s press release, not findings by the court.
Szpunar’s opinion draws a distinction between facts and expression. A chatbot does not infringe the press publishers’ right merely by reporting the same facts as an article, he said. The risk arises if the service obtained the article by bypassing an access restriction, such as a paywall or anti-scraping measure, or if the answer preserves the article’s distinctive wording or presentation so closely that users can dispense with the original. The opinion said that, on the facts available, there appeared to be no access barrier, no recognisable wording in the example answer, and safeguards at Google.
The Advocate General also recommended that the court reject the case’s two questions about AI training as inadmissible because the record did not establish that Like Company’s articles were used to train Gemini. He added, as a separate observation, that training is generally covered by the EU text-and-data-mining exception, subject to rights holders’ ability to opt out in a machine-readable form. That discussion is not the basis of his proposed answer to the dispute and does not establish that the articles were used for training.
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
When AI Summaries Cross the Line
The opinion could shape how publishers and AI companies assess live retrieval and summarisation of news. Its proposed test would not give publishers control over the underlying facts or every short summary. Instead, it focuses on how a chatbot obtains content and whether its answer substitutes for the article by retaining recognisable expression.
For services that search the web in response to a user’s prompt, the opinion points to practical risks around respecting access controls and avoiding answers that closely reproduce a source. For publishers, it suggests a possible basis to challenge particular uses or seek licensing discussions, but not a general right to prevent AI systems from conveying reported facts. The final legal effect will depend on the Court of Justice’s judgment and how the Hungarian court applies it.
Readers may also be affected: a ruling that discourages near-copies could lead to more paraphrased answers, source links, or summaries drawing on several publications. The opinion does not require any specific product changes, and it does not establish what Google or other chatbot providers will do.
Popular right now
De'Longhi Magnifica Duo Fully Automatic Espresso Machine with Milk Frother, Silver
LG 85-Inch QNED84B Mini LED 4K Smart TV (2026)
Ninja Foodi Air Fryer, 8 Qt DualZone, 6-in-1, Grey, DZ201
As an affiliate, we earn on qualifying purchases.
From News Aggregators to Chatbots
The EU created a related right for press publishers in Article 15 of the 2019 Copyright in the Digital Single Market Directive. As Szpunar recounts, the measure was intended in part to let publishers control and license certain uses of excerpts by online services such as news aggregators. The present case asks whether that right also applies when a chatbot retrieves an article and formulates a response to a user.
Like Company brought the dispute before a Hungarian court, which referred four questions to Luxembourg: two concerning training and two concerning chatbot answers. The Court of Justice’s press release says nine EU governments and the European Commission submitted written observations, and that the hearing took place on March 10, 2026. The court describes the case as the first generative AI case to reach it.
On training, Szpunar reasoned that the record did not support the claim that the specific dolphin article was part of Gemini’s training material. The article appeared in July 2023, while Gemini 1.0 was released in December 2023; he considered training on that later-published piece implausible. He said the example was instead consistent with retrieval-augmented generation, in which a system searches for material at the time of a request and uses it to compose an answer. His separate comments on the text-and-data-mining exception were made for completeness, not as a finding that training occurred in this case.
The Court Still Has to Rule
The opinion is an adviser’s recommendation, not a judgment, and the Court of Justice has not announced when it will issue its decision. The court may follow, modify or reject Szpunar’s reasoning. The final ruling will provide the court’s answer to the legal questions referred from Hungary, but the referring court will still need to resolve the underlying dispute.
Several factual matters also remain unresolved. The record described in the opinion did not establish that Like Company’s articles were used to train Gemini, and the case’s training questions may be dismissed on that basis. The Court of Justice has not determined whether Google infringed any rights, whether the example is representative of other chatbot answers, or how the proposed test would apply to different access controls and forms of reuse.
A Pending Luxembourg Judgment
The next milestone is the Court of Justice’s judgment in Case C‑250/25. No judgment date has been set. The court’s statistics cited in the source material show that preliminary references took an average of 16.9 months in 2025, but that figure does not predict the timing of this case.
After the EU court answers the referred questions, the Budapest court will apply that guidance to Like Company’s claims against Google. Until then, the opinion offers a proposed legal framework rather than a final determination of liability. The training issue, the status of the example summary and the wider effect on chatbot products remain subject to the courts’ decisions.
Key Questions
Has the EU Court of Justice ruled that Gemini infringed copyright?
No. Szpunar issued a non-binding opinion on November 3, 2026. The Court of Justice has not yet delivered its judgment, and the Hungarian court must decide the underlying dispute.
When might an AI chatbot summary infringe publishers’ rights?
Under the Advocate General’s proposed approach, the issue may arise if a chatbot gets around an access restriction or reproduces an article’s expression so recognisably that users can avoid the original. Reporting the same facts alone would not meet that test.
Did the opinion find that Google trained Gemini on Like Company’s articles?
No. Szpunar said the record did not support that claim and recommended that the training questions be treated as inadmissible. His separate comments about text-and-data mining were not a finding that the articles were used for training.
What happens after the Advocate General’s opinion?
The Court of Justice will issue a judgment at a date it has not yet announced. The Budapest court will then apply the EU court’s guidance to the case between Like Company and Google.
Source: ThorstenMeyerAI.com
