Maciej Szpunar, First Advocate General at the Court of Justice of the European Union (CJEU) said on November 3 that Google’s Gemini AI chatbot may violate European copyright rules to answer the users. The statement could affect how technology companies use news and how publishers protect their work.
Advocate General Maciej Szpunar presented his opinion in a case involving Google and Like Company(A Hungarian digital news publisher). Like Company says Gemini used its articles to answer user questions without permission.
He argued that a chatbot may violate a publisher’s rights when it uses an online article to create a response and then shares it with a user. The risk depends on how the chatbot obtains the article and how much of the original content it includes in its answer.
However, copyright law does not protect basic facts or give publishers control over them. News organizations cannot stop others from reporting events such as a company launching a new product or a government introducing a policy. Instead, the law protects how a writer presents those facts not the facts themselves.
He clearly separated the use of text to train AI systems from the answers they provide to users. Szpunar said the court should not consider the training issue because there was no evidence that Like Company’s articles were used to train Gemini.
He also pointed out that EU law generally allows AI companies to use copyrighted material for training under its text and data mining rules. However, copyright owners can prevent such use by clearly stating that they do not allow their content to be used for this purpose.
Basically, every AI model learn from large amounts of text and use that knowledge to generate answers, explain subjects and summarize information. Some AI chatbots can also search the internet for latest information. Google Gemini uses a method called retrieval-augmented generation (RAG) to find online content and include relevant information to provide answers.
The case is the first generative AI case to reach the EU court in this field. The case has raised concerns among news publishers that rely on website traffic, subscriptions and advertising for income. When AI chatbots provide complete answers or detailed summaries of news stories, readers may stop visiting the original news websites. This could reduce website traffic, cut advertising income and make it harder for publishers to attract new subscribers.
EU laws give news publishers certain rights over how their content is used online. The 2019 Copyright Directive also set some limits exceptions, including rules that allow text and data mining in certain situations. At the same time, it does not stop people from linking to news articles or using small parts of their content.
Szpunar’s opinion does not require Google to pay publishers or resolve the dispute. The CJEU judges will examine the case and deliver the final decision, which will be legally binding. The court often follows an Advocate General’s opinion, but it is not required to agree with it.
The final decision could affect how AI companies use news content, how chatbots share information from articles and whether publishers can demand payment for using their work across the European Union. It may also set clearer limits for the growing use of AI in online news.
