A German court has ruled that Google is directly liable for AI-generated content shown to users through its AI Overviews feature on its search platform.
The Regional Court of Munich recently issued a temporary injunction barring Google from spreading false claims about two Munich-based publishers through AI Overviews answers. The ruling stems from a lawsuit filed by the publishers against Google over errors in AI Overviews answers that described them as scam websites even though they are genuine businesses.
The court also noted that the errors were not found in the websites that had been cited as sources in the AI-generated summary of search results provided by AI Overviews.
Although the ruling is preliminary, it has gained significant attention as it could have broader implications for the AI industry. It raises a key unresolved question about who is responsible for costly factual errors or hallucinations by AI models. Should liability be borne by the model provider, the platform that hosts AI-generated content, or the user who submits such prompts?
The ruling also comes at a time when debates are intensifying globally over whether safe harbour provisions – that give online platforms legal immunity from content that appears on their services – are still appropriate, given their role in enabling the spread of misinformation, riots, and even acts of terrorism.
In India, a similar debate emerged earlier this year around the Grok AI chatbot developed by Elon Musk-owned SpaceXAI (formerly xAI). GPT-based models essentially generate responses by predicting the next token in a sentence or the next pixel in an image based on patterns learned from vast training datasets. But the models and the data they are trained on are ultimately products of humans.
That is what makes assigning accountability tricky. Many argue that AI companies should be held responsible for the outputs of their systems. The German court’s ruling firmly aligns with that view. “These are unique assertions invented by Google’s AI tool” which Google “must accept responsibility for,” the ruling read.
“We invest deeply in the quality of AI Overviews to ensure that the overwhelming majority of responses provide accurate information, and they are designed to reflect the information that exists on the web. We’re carefully reviewing this decision, which is not yet final,” a Google spokesperson was quoted as saying by Android Authority in response to the ruling.
What was the lawsuit about?
Two publishing companies filed a lawsuit against Google, alleging that its AI Overviews answers falsely tied them to scams, subscription traps, and shady business practices for certain search queries.
The names of the publishers have been redacted in the 35-page ruling dated May 28, 2026. The court said in its ruling that AI Overviews had mixed up information about other, genuinely sketchy companies with the plaintiffs and drew connections that did not appear in any of the linked sources, according to a report by Decoder.
The plaintiffs said that they had sent Google a cease-and-desist letter, but the company did not respond appropriately, according to the court.
What did the court say in its ruling?
The German court ruling classifies Google as a direct infringer since AI Overview-generated content is its own content, and not just a list of blue links shown as search results. The key findings of the court include the following:
-Search engine liability: Typically, tech firms such as Google enjoy legal protections under safe harbours laws that vary from region to region. However, the court ruled that previous case law shielding search engine operators from liability does not apply to AI Overviews.
“For standard search engines, courts usually rule that operators do not have to check every link in advance,” the court said after examining previous rulings from Germany’s Federal Court of Justice (BGH), which gave traditional search engines limited liability. This ruling had argued that search engine operators were only liable as indirect infringers because they merely made third-party content findable.
However, the Munich court found that this reasoning does not apply to AI Overviews because it generates “independent, new, and substantive statements” by evaluating and combining content from various third-party sites as opposed to regular search engines that just point to outside websites.
-AI Overviews versus traditional search results: In its ruling, the court said that Google’s AI Overviews do not work like traditional search results as it “weaves information together to create brand-new text statements.” It pointed out that the AI Overviews answers highlighted by the plaintiffs open with confident claims like “Yes, [company] is known for dubious business practices,” and then builds its own structure with a summary, red flags for the alleged scam, and tips for users.
It also held that AI Overviews make claims “that are not even made in the search results.” Referring to the allegations in the lawsuit, the court said that since none of the linked sources drew any connection between the plaintiffs and the shady companies mentioned in the AI Overviews, these answers are “the defendant’s own statements.”
As a result, Google “alone has influence over the AI’s offering and the algorithms with which the AI operates,” it said.
-Accuracy of AI Overviews: During hearings, Google argued that users could verify AI-generated answers by looking at the websites linked as sources in the summary. It also said that users “generally knew “that information generated with AI should not be blindly trusted.
However, the court rejected Google’s arguments and said, “If an AI Overview were legally treated as completely unreliable text that required users to check every single link anyway, the entire feature would lose its stated purpose.”
The AI Overview contained “a self-contained statement with independently understandable content and no reference to other possible interpretations or even unreliable content,” it said. Citing press law as an example, the court said that publishers are liable for teasers that are understandable on their own, even if readers never read the full article.
-Regulatory gap: The court reportedly pointed out that if Google were only liable for obvious violations, victims would have no real legal recourse when the AI makes false claims. Since the third-party websites cited as sources had not even made the false statements in question, the victims could not sue them. Under existing rules, they cannot effectively sue Google either.
As a result, Google cannot invoke host provider protections under the EU’s Digital Services Act or fall back on the standard notice-and-take-down process for search engines, the court reportedly said.
-Free speech protection for AI-generated content: In its ruling, the court said that an AI-generated opinion is “not the expression of an acquired conviction of the persons expressing it, but the result of an algorithm.” AI-powered search is “above all an expression of Google’s business activities” and “at most a secondary expression of an interest in being able to freely express one’s opinion and beliefs.”
What are the remedies imposed on Google?
The court ruling is in favour of the plaintiffs on most counts, except for two minor requests. The temporary injunction bans Google from making claims about the plaintiffs related to scams, connections to dubious companies, subscription traps, phone calls that never happened, and lack of availability through its AI Overviews offering.
Additionally, Google has been directed to cover 80 per cent of the legal costs of the plaintiffs, with each of them paying 10 per cent each.
Sundar Pichai, the chief executive of Google. (Mike Kai Chen/The New York Times) 
Share
