Google to appeal German ruling on AI Overviews liability

A German ruling has forced a blunt question onto the AI industry: when a generative search summary invents defamatory facts, can the platform be held responsible for publishing them? Google said it will appeal after the Munich Regional Court ordered it to stop repeating challenged statements in AI Overviews, a decision that could influence how search engines and AI assistants are designed, moderated and litigated across Europe.

The case was brought by two German publishers who said Google’s summaries falsely linked them to scams and dubious business practices. In summary proceedings on May 28, 2026, in case 26 O 869/26, the court’s 26th civil chamber, which specializes in press and defamation law, treated the AI-generated text as Google’s own content rather than as a neutral index of third-party material. The court also rejected Google’s argument that users could simply verify the answer by clicking cited links, finding instead that the AI response appeared as a self-contained statement with independently comprehensible content.

The ruling adds immediate legal and commercial pressure to a feature that now sits above traditional search results and is often the first thing users see. If appellate judges uphold the decision, companies that generate summaries from web content could face broader liability when those systems make factual accusations about people, publishers or businesses. Reporting cited JUVE as saying Google may have to bear about 80 percent of the legal costs, underscoring how expensive these disputes can become even before a wider test reaches the European Union.

Google said the case involved specific errors rather than any flaw in the underlying product. The company also said the vast majority of AI Overviews are accurate, while acknowledging that, like any search feature, they can sometimes miss context or misread source material. That defense will now be tested against a legal theory that increasingly treats AI output not as a passive display of the web, but as a publisher-like act with real reputational consequences.

The Munich decision did not emerge in isolation. It built on a Frankfurt am Main Regional Court ruling from September 2025 that said liability for false AI overviews is not automatically excluded. At the same time, regulators are already circling the issue on the competition side. European publishers filed an antitrust complaint with the European Commission in July 2025, and German media groups filed another with Germany’s Federal Network Agency in September 2025, arguing that AI summaries can drain traffic, readership and revenue by answering searches before users click through.