A Munich regional court ruled in late May that Google can be held directly responsible when its Google AI Overviews feature publishes false statements in the top-of-search summaries that are not supported by the cited sources. The case was brought by two Munich-based publishers; the plaintiffs' names were not disclosed in the public court filings (for example, file number 26 O 869/26). Reuters reported that Google intends to appeal the decision.
The facts of the case
The plaintiffs said that when searching for their companies, Google AI Overviews produced misleading negative claims linking them to scams and abuses that did not appear in the original source posts. The court found that if statements in the overview do not follow from the used sources but are instead produced by the AI, the overview counts as content created by Google itself — making Google directly responsible for it.
How Overviews works and the related risks
Google Overviews provides a bullet-point summary based on available links in response to a search query, instead of or in addition to a list of links. While the underlying links are available, practical experience shows users often read only the AI-generated summary and seldom click through to source articles, as noted by Bartal Iván, partner at Oppenheim Ügyvédi Iroda.
Generative language models are known to 'hallucinate' at times — they can generate assertions that are not present in the source material. In this lawsuit, the court found exactly that: the negative claims did not appear in the original postings, so the summary was deemed Google-generated content.
Legal background and previous practice
Historically, legal regimes — initially in the United States about 30 years ago and later adopted in European regulation — provided platforms and intermediaries certain protections from liability for third-party content, provided they only stored or transmitted links and removed unlawful content when notified. This framework helped platforms such as Google, YouTube, Facebook, X, and others to grow into dominant global players.
The Munich ruling, however, draws a line where a service actively creates or transforms content within its own interface: at that point the provider can no longer claim passive intermediary status to avoid liability.
Consequences and remedies
Under the court's reasoning, victims can seek more than just removal of a link. Within the EU legal framework, affected parties may claim non-pecuniary damages (sérelemdíj), seek a declaration that the right was violated, obtain injunctions to stop the unlawful conduct, and demand corrective statements. If they can prove material harm from false statements in the Overviews — for example loss of clients or employment — they may also claim pecuniary damages and legal costs.
Potential impact in Hungary
Because the issue is regulated at EU level, a similar case could be brought before a Hungarian court. Hungarian law, supplemented by EU rules, also allows direct action against a foreign service provider and the pursuit of damages if the legal conditions are met.
What happens next for Google?
Industry data cited in the reporting suggest Google AI Overviews operates at roughly 90 percent accuracy. Nevertheless, the feature's ubiquity and Google's market dominance create substantial financial and liability exposure, amplified by differing national rules on damages. Google has announced it will appeal the Munich decision. It remains to be seen whether Alphabet will change the product's operation or adopt other risk-mitigation measures; the company is more likely to pursue appeals and iterative product adjustments rather than scale back the summaries' assertiveness.
The ruling may set an important precedent within the EU for when a platform-generated summary is considered the provider's own content and when the provider can be held directly responsible for the statements it contains.
