🔍 Read the full analysis: A Closer Look At The EU View On Chatbots And Publishers’ Rights 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 retrieving and reusing news articles may infringe publishers’ rights in certain circumstances. He said the test appears not to be met in the Gemini case involving a Hungarian article about dolphins; the opinion is nonbinding, and the Hungarian court will decide the dispute.
The European Union’s top court has received its first generative AI case, after Advocate General Maciej Szpunar advised that chatbots retrieving and reusing news articles may infringe press publishers’ rights in some circumstances. In an opinion issued Nov. 3, 2026, he said infringement could arise if a service accessed an article unlawfully or reproduced its distinctive expression so extensively that users could skip the original, but his proposed test appears not to be met in the dispute over Google Gemini and a Hungarian article about dolphins.
The case, C‑250/25, Like Company v Google Ireland, concerns a Hungarian publisher that says Google used its articles in Gemini answers and in training the chatbot’s models. Like Company runs several Hungarian news portals. According to the Court of Justice’s press release, it alleges Google repeatedly reproduced and made its articles available through Gemini from June 13, 2023, to May 31, 2024, as well as using them to train language models. Those are the publisher’s claims, not findings by the court.
The example at the centre of the dispute was a Hungarian article published in July 2023 about a singer’s reported plan to bring dolphins to Lake Balaton. A user asked Gemini to summarise the article, and the chatbot produced a Hungarian-language summary. Szpunar’s opinion says answers may infringe the publishers’ right if the chatbot obtained the material by circumventing access restrictions, or if the article’s particular wording and presentation are recognisable in the response to the point that readers can forgo the original publication.
But facts themselves are not protected by the press publishers’ right, the opinion says; the protected element is the article’s expression. Szpunar concluded that the alleged conditions do not appear to be present in this case: he found no indication of access barriers being bypassed or recognisable wording in the example answer, and noted that Google appears to have safeguards. These are the Advocate General’s assessments, not a final ruling. The Hungarian court referred four questions to Luxembourg, including two concerning 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
The Line Between Summaries and Copying
The opinion offers a proposed boundary for companies that build chatbots or AI search tools using material retrieved from the web. Under Szpunar’s approach, a system could face a publishers’ rights issue if it evades a paywall or anti-scraping measure, or if its answer closely preserves a publication’s distinctive expression rather than merely conveying the same facts. The test focuses on how the material was accessed and what the answer reproduces.
For publishers, the approach could support challenges to certain chatbot outputs and strengthen arguments for licensing where answers substitute for reading an article. It does not, however, give publishers control over facts reported in their stories, nor does it establish that every AI summary requires permission. For users, a future ruling could affect how chatbots quote, summarise and link to news reporting. The practical result may depend on the wording of answers, the sources used and the protections in place on publishers’ sites.
The opinion may inform a broader debate about the relationship between AI services and news organisations, but its immediate legal effect is limited: it is advice to the court, not a judgment. The Court of Justice’s eventual ruling, followed by the Hungarian court’s application of the law to the evidence, will determine the case.
Products Worth Considering
Logitech MX Mechanical Wireless Illuminated Keyboard, Tactile Quiet, Graphite
Canon SELPHY CP1500 Compact Photo Printer (White)
Apple MacBook Pro 14 2021 with M1 Max, 64GB RAM, 2TB SSD, Silver
As an affiliate, we earn on qualifying purchases.
How the Gemini Dispute Reached Luxembourg
EU law has protected a press publishers’ right since the 2019 Copyright in the Digital Single Market Directive. The provision at issue, Article 15(1), gives publishers rights over certain online uses of their press publications. The Advocate General’s opinion recalls that legislators adopted the right with services such as news aggregators in mind. The current case asks how that framework applies when a chatbot retrieves an article and generates an answer for a user.
The Budapest-area court sent four questions to the Court of Justice. Nine EU governments and the European Commission submitted written observations, and the hearing took place on March 10, 2026. Szpunar treated the training questions as inadmissible because, in his view, the record did not establish that Like Company’s articles were used to train Gemini. He noted that the dolphin article appeared in July 2023, while Gemini 1.0 was released in December 2023, and considered it implausible that the article had been part of the model’s training data.
Instead, he said the example was consistent with retrieval-augmented generation: the chatbot searches for material when prompted, retrieves pages and uses them to produce an answer. The user had included the publisher’s domain in the prompt. Szpunar also reasoned that training on dedicated computers is not the “online use” covered by the publishers’ right. Separately, and as additional analysis rather than a ruling on the case, he said model 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.
What the Opinion Does Not Decide
The Advocate General’s opinion is not binding, and the Court of Justice has not yet issued its judgment. No date for that ruling has been set. The opinion’s conclusion that the proposed infringement conditions do not appear to be met is provisional and based on the material described in the reference; the Hungarian court must decide the dispute using the evidence before it.
It also remains unresolved how the Court of Justice will address the questions referred by the Hungarian court, including the legal framework for chatbot answers and the treatment of AI training. Szpunar advised that the training questions were inadmissible because the case record did not support them. His additional discussion of text and data mining was offered for completeness, not as the Court’s decision in this case. The full impact on other chatbot services will depend on the final judgment and the facts of future disputes.
Awaiting the Court’s Judgment
The Court of Justice will now consider the Advocate General’s opinion before issuing its ruling. The opinion is advisory and does not bind the judges, and no judgment date has been announced. The court’s statistics put the average duration of preliminary-reference cases at 16.9 months in 2025, but that figure is historical and does not establish when this case will be decided.
After the EU court answers the legal questions, the Hungarian court will resume the case and apply the ruling to Like Company’s claims against Google. Until then, the opinion supplies a proposed test, not a final determination that Gemini infringed publishers’ rights or that the challenged use was lawful.
Key Questions
Did the EU Court of Justice rule that Gemini infringed copyright?
No. An Advocate General issued a nonbinding opinion on Nov. 3, 2026. The Court of Justice has yet to give its judgment, and the Hungarian court will decide the dispute.
When could a chatbot summary infringe a publisher’s rights under the opinion?
Szpunar said infringement may arise if a chatbot accessed an article unlawfully, such as by bypassing a paywall or anti-scraping block, or if it reproduced the article’s distinctive expression so extensively that users could skip the original. Repeating the same facts alone is not enough under his analysis.
Did Szpunar find that Gemini trained on Like Company’s articles?
No. He said the case record did not show that the articles were used in training and regarded that possibility as implausible for the example article. He treated the training questions as inadmissible, then discussed text-and-data mining rules separately for completeness.
What happens next in the case?
The Court of Justice will issue a judgment after considering the opinion; no date has been announced. The Hungarian court will then apply the EU court’s answers to the case evidence.
Source: ThorstenMeyerAI.com
