German courts treat Google AI search answers as Google’s own statements

Two German regional courts have now treated Google’s AI-generated search answers as statements Google can be liable for, according to The Decoder. The latest preliminary injunction, from Aachen, matters because it applies the same logic as an earlier Munich case even though Google settled that appeal before a higher court could rule.

Aachen follows Munich on AI search liability

The Aachen Regional Court has barred Google from spreading specific false claims about a doctor through its AI search results, The Decoder reports. The court relied on reasoning from a late-May ruling by Munich Regional Court I, which held that Google’s AI-generated search summaries are not merely links or neutral references to third-party pages.

In the Munich case, Google’s AI summaries had falsely connected two Munich-based publishers with fraud schemes. The judges treated those AI Overviews as Google’s own statements, meaning the company itself could be responsible for unlawful content generated by the feature.

Google announced an appeal in that case, but The Decoder says the appeal never produced a higher-court decision because the matter ended in a settlement. The reasons for the settlement, and whether money was involved, are unknown. The Decoder says neither Google nor the affected publisher commented, and the court did not address the broader significance beyond the individual dispute.

That makes the Aachen decision important. It does not create a final nationwide precedent, and Google can still challenge the injunction. But it shows another German court adopting the same legal theory: when Google’s AI combines and evaluates material from third-party sites to produce a new answer, the resulting text may be treated as Google’s own content.

The doctor case shows the reputational risk

The Aachen dispute involved a physician who had worked as chief physician and professor of neuroradiology from 2010 until late June 2026. On August 28, a colleague told him that Google’s AI Mode was explaining his departure by referring to an alleged sexual assault involving his secretary.

According to The Decoder, follow-up searches using terms such as suspension, termination and court proceedings produced detailed multi-paragraph fabrications. The system also invented a legal aftermath. The allegations did not relate to the doctor at all; the court found that the AI had assigned events involving another former university professor to him.

The doctor emailed Google in late August, described the error and asked for the content to be removed. The company said in mid-September that it would take no action, according to the court account cited by The Decoder. A formal legal demand also did not resolve the issue, and the court said the false AI-generated text remained live.

The Aachen court then issued an emergency order without first hearing Google, citing the potential damage to the doctor’s professional reputation. If Google violates the order, it faces fines of up to 250,000 euros per violation.

The court also said Google could check AI outputs, for example by comparing them with the source websites used to generate them. The judges added that if Google had confronted its own AI system with the doctor’s objections, the system would have detected the errors itself.

Why businesses and creators should pay attention

For small businesses, freelancers and professional creators, the practical issue is not only whether an AI answer is useful. It is whether an AI search product can confidently attach a damaging claim to the wrong person or company, and whether the platform has a process to fix it quickly.

The rulings concern Google’s AI search features, not the standalone Google Gemini chatbot. Still, they sit inside a broader shift toward AI systems that summarize information for users rather than sending them directly to source pages.

That shift changes the risk profile for anyone whose reputation depends on search results. A traditional blue link may point users to a third-party page, but an AI-generated paragraph can appear to state a conclusion in the search product’s own voice. The German courts’ reasoning turns that product design choice into a liability question.

For readers who manage brands, clients or public profiles, the takeaway is practical:

• Monitor AI search answers for your own name, business name and key executives.
• Pay attention to queries that combine your name with sensitive terms.
• Keep records of false outputs and the searches that produced them.
• Treat AI search errors as reputational incidents, not just technical glitches.

The unresolved publisher question

The Decoder notes that the legal issue could extend beyond defamatory or false claims. If AI answers are treated as Google’s own content, publishers whose material is used to generate those answers might eventually bring their own claims. That was not part of the Munich case, but The Decoder frames it as a possible next step if courts continue down this path.

That connects with a wider debate over how AI systems use publisher material, including questions Emonarc has covered around AI training and publisher websites. The German rulings do not decide that issue, but they make the boundary between search reference and generated content harder to ignore.

For Google, the immediate question is whether it challenges the Aachen injunction and risks a more binding decision, or continues resolving cases individually. For users, the lesson is simpler: AI search answers may look authoritative, but courts are now being asked who is accountable when they are wrong.