🔍 Read the full analysis: Can Chatbots Reuse Facts But Not Publishers’ Writing? 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
On Nov. 3, 2026, Advocate General Maciej Szpunar advised the EU Court of Justice that chatbots may infringe press publishers’ rights when they use articles unlawfully or reproduce their recognizable expression. He said the test does not appear to be met in the dispute over Gemini’s summary of a Hungarian article, and the opinion is not binding.
Advocate General Maciej Szpunar advised the European Union’s top court on Nov. 3 that chatbots can infringe press publishers’ rights when they retrieve and reuse news articles in answers, but only under specific conditions. In Case C‑250/25, involving Google’s Gemini chatbot and a Hungarian publisher, he said infringement may arise if a system obtains an article by bypassing access restrictions or reproduces its distinctive wording so extensively that readers can avoid the original publication.
The opinion draws a distinction between facts and expression: reporting information from an article is not, by itself, enough to infringe the EU press publishers’ right. The answer may cross the line if the chatbot’s wording makes the article’s original form recognizable and lets users dispense with reading the source, or if the system accessed the article unlawfully, for example by circumventing a paywall or anti-scraping measure. Szpunar described this as a case-specific test, not a general ban on chatbots summarizing news.
The dispute began after Like Company, a Hungarian publisher whose outlets include news portals, sued Google. The company said Gemini repeatedly reproduced and made its articles available to the public between June 13, 2023, and May 31, 2024, and also alleged that its articles were used to train Google’s language models. One example concerned a July 2023 article about a singer’s reported proposal to bring dolphins to a pool on Hungary’s Lake Balaton peninsula of Tihany. Gemini produced a summary in Hungarian after a user asked about the article.
Szpunar’s assessment of that example points against the publisher’s claim under the test he proposed. He said the material before him did not indicate that Google bypassed access restrictions, and the example answer did not appear to reproduce recognizable wording from the article. He also noted that Google appeared to have safeguards. The Hungarian court will decide the case; the Advocate General’s opinion is advice to the Court of Justice, not a binding judgment.
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 Line for News Reuse
The proposed test matters to both publishers and AI services because it focuses on how a chatbot obtains and presents material, rather than treating every reuse of reported information as a rights violation. Publishers could have grounds to challenge answers that evade technical or paid access controls, or that function as substitutes for reading an article. But they would not, under this approach, gain exclusive control over the underlying facts simply because those facts first appeared in their reporting.
For chatbot operators, the opinion points to practical risks around access controls and close paraphrase. A system that searches the web while answering a prompt may need to account for whether it was permitted to retrieve a page and whether its output reproduces the source’s distinctive expression. For users, the dispute could affect how AI search tools summarize news and whether they direct readers to original reporting. These are potential implications of the proposed legal test, not changes already imposed by a final court ruling.
The outcome could also influence licensing discussions. The EU press publishers’ right, introduced in 2019, was designed in part to give publishers control over uses of news extracts by online services such as news aggregators. Szpunar’s opinion addresses whether that right can apply to chatbot answers, while leaving the question to the Court of Justice to settle.
Popular right now
Sony BRAVIA XR8B 65-Inch 4K OLED Google TV (K-65XR8B)
De'Longhi Dinamica Plus Fully Automatic Espresso Machine
eufy C28 Robot Vacuum Cleaner and Mop for Pet Hair, 15,000 Pa
As an affiliate, we earn on qualifying purchases.
How the Gemini Dispute Reached Luxembourg
Like Company brought its case in Hungary, arguing that Google had infringed the press publishers’ right established by Article 15 of the 2019 Copyright in the Digital Single Market Directive. The Budapest regional court referred four questions to the Court of Justice: two on the alleged use of articles for AI training and two on their use in chatbot answers. The hearing took place on March 10, 2026, after written observations from nine EU member states and the European Commission.
Szpunar treated the training questions differently from the chatbot-answer questions. He advised the Court to find the training questions inadmissible because the record did not support the claim that Like Company’s articles were used to train Gemini. The dolphin article appeared in July 2023, while Gemini 1.0 was released in December that year. The opinion says the timing made training on that particular article implausible and points instead to live retrieval from the web as the likely explanation for the summary.
Only as an additional observation, Szpunar discussed the EU’s text and data mining exception. He said AI training is, in principle, covered by that exception, while rights holders can opt out in a machine-readable form. That discussion was not his main answer to the case: he first concluded that the training questions should not be decided because the necessary factual basis was missing.
The Court Has Yet to Rule
The Advocate General’s opinion is not binding, and the Court of Justice may adopt a different approach in its eventual judgment. No date has been announced for that ruling. The Court’s questions concern the interpretation of EU law; the Hungarian court must apply the eventual answer to the evidence in the dispute.
Several facts also remain unresolved. The source material does not establish that Like Company’s articles were included in Gemini’s training data, and Szpunar advised that the related questions should be dismissed as hypothetical. His separate comments on the text and data mining exception were made for completeness, not as a ruling that resolves Google’s conduct. The precise reach of the proposed test across different chatbot designs, access arrangements and types of output will depend on the Court’s judgment and later cases.
Awaiting the Court’s Judgment
The next major step is the Court of Justice’s judgment in Case C‑250/25. The court has not given a date. Its ruling will answer the Hungarian court’s questions about the EU press publishers’ right and chatbot responses; the Hungarian proceedings can then continue with that interpretation in hand.
Until then, the opinion offers guidance rather than a final legal outcome. Publishers, AI companies and users will be watching for whether the Court adopts Szpunar’s distinction between freely reusable facts and protected expression, and how it treats access restrictions and summaries that may replace the original article.
Key Questions
Has the EU Court of Justice ruled that Gemini infringed the publisher’s rights?
No. Advocate General Maciej Szpunar issued a non-binding opinion. He said the conditions for infringement did not appear to be met in the example before him, but the Court has not yet issued its judgment.
When could a chatbot answer infringe a publisher’s rights under the opinion?
Szpunar said infringement may arise if a chatbot obtains an article by bypassing access restrictions or reproduces its distinctive expression so recognizably that users can avoid accessing the original publication. Reusing the same facts alone would not be enough under the test he proposed.
Did the opinion establish that Google trained Gemini on Like Company’s articles?
No. Szpunar advised that the training questions were inadmissible because the record did not show that the articles were used for training. He considered live retrieval from the web a likely explanation for the example summary.
What did the opinion say about AI training and copyright?
After advising that the training questions should not be decided, Szpunar discussed the issue for completeness. He said the EU text and data mining exception covers AI training in principle, with rights holders able to opt out in a machine-readable way. That discussion is part of a non-binding opinion, not a final ruling in this case.
What happens next in the case?
The Court of Justice will issue a judgment answering the Hungarian court’s questions, but it has not announced a date. The Hungarian court will then apply the EU court’s interpretation to the dispute between Like Company and Google.
Source: ThorstenMeyerAI.com
