🔍 Read the full analysis: AI Chatbots And Copyright: Where Facts End And Publishers’ Writing Begins 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 November 3, 2026, Advocate General Maciej Szpunar advised the EU Court of Justice that chatbot answers using news articles may infringe press publishers’ rights in limited circumstances. His proposed test focuses on unlawful access or reuse of recognisable expression, not the repetition of facts; the opinion is nonbinding, and the case concerns a Gemini summary of a Hungarian article about dolphins.
Advocate General Maciej Szpunar advised the European Union’s Court of Justice on November 3 that AI chatbots can infringe press publishers’ rights when they retrieve and reuse news articles in answers, but only in specific circumstances. In Case C‑250/25, Like Company v Google Ireland, he said the key questions are whether the service accessed the article unlawfully or reproduced its distinctive expression so closely that users could avoid reading the original.
The dispute began after Google’s Gemini chatbot produced a Hungarian summary of a July 2023 article about a singer’s reported plan to bring dolphins to Lake Balaton. Hungarian publisher Like Company, which operates several news websites, sued Google. It argued that the company used its articles both to train Gemini and to generate answers that reproduced and made the material available to users. Those are allegations in the case, not findings by the EU court.
Szpunar’s opinion says that a chatbot may infringe the press publishers’ right if it gets material by circumventing a paywall or an anti-scraping measure, or if its answer preserves recognisable expression from an article to the point that readers can dispense with the original publication. Repeating facts alone is not enough, under the test he proposed. The opinion points to the difference between free-to-report information and the protected way a journalist expresses it.
On the example before the court, Szpunar said the conditions for infringement did not appear to be met: the article was not protected by an access barrier, the sample answer did not appear to reproduce recognisable wording, and Google appeared to have safeguards. The Hungarian court hearing the dispute must decide how the legal test applies to the evidence. The Advocate General’s opinion is not a final ruling and does not determine whether Google is liable.
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 Test Turns on Access and Expression
The proposed approach could shape how publishers and AI companies negotiate over news content. It gives publishers a possible claim when a chatbot bypasses technical restrictions or delivers an answer that functions like a substitute for an article. But it does not give publishers control over the underlying news facts or every summary that describes them.
For chatbot and AI-search providers, the opinion puts attention on how content is retrieved and on whether generated answers reproduce a source’s distinctive wording or structure. For users, the outcome could affect how services balance short answers, source links and summaries drawn from multiple publications. Those are possible implications of the proposed test, not rules already settled by a court judgment.
Publishers’ rights under EU law have applied since 2019, but the law was adopted with online uses such as news aggregation in view. The case asks how that framework applies to a service that retrieves articles and uses them to answer individual questions. The answer may affect licensing discussions, but the opinion does not establish a general right to charge for the reporting of facts.
Products Worth Considering
Logitech MX Mechanical Wireless Illuminated Keyboard, Tactile Quiet, Graphite
Dell SE2426HG 24-Inch FHD 240Hz Gaming Monitor
Logitech MX Brio Ultra HD 4K Webcam, Graphite
As an affiliate, we earn on qualifying purchases.
How the Gemini Dispute Reached Luxembourg
Like Company brought its case in Hungary, and the Budapest regional court referred four questions to the EU Court of Justice: two about AI training and two about chatbot answers. The hearing took place on March 10, 2026. The court’s press release said Like Company alleged that Google used its articles to train Gemini and repeatedly reproduced them through the chatbot between June 13, 2023, and May 31, 2024. Nine EU governments and the European Commission submitted written observations.
Szpunar advised the court to treat the questions about training as inadmissible because the case record did not establish that the article at issue was used in Gemini’s training data. He reasoned that Gemini could have summarised it by retrieving the page live through retrieval-augmented generation, a method in which a system searches for relevant material and generates an answer based on what it finds.
He also discussed training separately, as an additional point rather than a decision on the case. In his view, training may in principle fall under the EU’s text-and-data-mining exception, subject to rights holders’ ability to opt out in a machine-readable way. That discussion does not amount to a binding ruling that any particular training practice is lawful.
“the first generative AI case to reach the Court of Justice”
— EU Court of Justice press release
The Final Judgment Is Still Pending
The Court of Justice has not yet ruled, and the Advocate General’s opinion is advisory rather than binding. The court may follow it, modify its reasoning or reach a different conclusion. No date for the judgment has been announced.
The Hungarian proceedings will also need to resolve the disputed facts, including what content Google retrieved, how Gemini generated its answer and what safeguards were in place. The opinion’s view that the training questions lack a factual basis is not a final finding about Gemini’s datasets or Google’s conduct in other cases.
The legal treatment of training remains less settled than the opinion’s proposed test for chatbot answers. Szpunar’s remarks on the text-and-data-mining exception were made for completeness after he advised that the training questions should not be considered on the available record. They should not be read as a final court decision covering all AI training or all rights-holder opt-outs.
Court Ruling and Hungarian Proceedings
The Court of Justice will issue its judgment in response to the Hungarian court’s referral; it has not announced when. The court’s statistics 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 clarifies the interpretation of the law, the Hungarian court will apply that interpretation to the dispute between Like Company and Google. Until then, publishers, AI providers and users have an advisory opinion proposing where the line may fall, but no final EU judgment in this case.
Key Questions
Has the EU Court of Justice ruled that Gemini infringed copyright?
No. Advocate General Szpunar issued a nonbinding opinion. The Court of Justice has not delivered its judgment, and the Hungarian court must decide the underlying dispute.
When might a chatbot answer infringe publishers’ rights under the opinion?
Szpunar proposed that infringement may arise if a service bypasses an access restriction, such as a paywall or anti-scraping measure, or reproduces recognisable expression so closely that users can avoid the original. Repeating facts alone would not meet that proposed test.
Did the opinion decide whether AI training on news articles is lawful?
No final ruling on training was made. Szpunar advised that the training questions should be inadmissible because the record did not show that the article was used for training. His separate discussion of the text-and-data-mining exception was not a binding decision on all training practices.
What happens after the Advocate General’s opinion?
The Court of Justice will issue a judgment, with no date set yet. The Hungarian court will then apply the EU court’s interpretation to Like Company’s case against Google.
Source: ThorstenMeyerAI.com
