Landmark German Court Ruling Holds Google Liable for AI-Generated False Claims, Setting Global Precedent

A regional court in Germany has issued a landmark ruling that could reshape how search engines and AI-powered chatbots operate across the world. The Munich Regional Court ruled in a preliminary decision that Google is legally responsible for a series of false statements generated by its AI Overviews feature, ordering the company to block the spread of erroneous or inaccurate content through its search platform.

The case originated from a report first published by Decoder, where two publishers discovered that Google’s AI-generated search summaries wrongfully tied their businesses to questionable practices, scams, and subscription fraud in specific search queries, with no factual basis for the connections. According to the report, the harmed companies sent Google a formal cease-and-desist letter earlier this year. Google rejected all liability, arguing that its automated summary feature already warns users AI content may contain errors and should be independently verified.

The court’s analysis found that Google’s AI mixed data on the plaintiff publishers with information about other companies already flagged for potential illegal activity, creating false associations that do not exist in any of the original sources linked by the search engine. The court noted that unlike traditional search engines, which only display lists of links to content created by third parties, Google’s AI tool produces “independent, new, and substantial statements” by misinterpreting information available online.

Per the court’s ruling, correcting misinformation is not the responsibility of outside third parties. Google is the only entity that can modify the underlying technology powering its AI-generated summaries, and therefore “must be held accountable” for harmful output. The court also rejected Google’s core defense, finding it lacked merit, because the challenged summary contained claims that do not appear anywhere in Google’s own linked search results.

---

A New (and Forceful) Interpretation of AI on the Web

The court’s framing of AI’s role in delivering search results positions this case as a historic legal precedent. It marks a major shift by holding a large tech company responsible for the output of its most advanced AI tools on widely used public platforms.

Until now, most global legal systems have categorized search engines as simple intermediaries that only facilitate access to third-party content hosted on the web. This status has long granted search providers a layer of legal protection when content linked through their platforms turns out to be false, inaccurate, misleading, or even defamatory.

However, the German court ruled that this existing legal protection no longer applies when search engines integrate generative AI systems. Per the court’s reasoning, generative AI is capable of fabricating entirely new claims by mixing content from multiple sources, so companies that operate these systems must assume liability for the content their AI produces.

Judges also concluded that while Google encourages users to verify AI content due to the inherent risk of hallucinations in AI models, this standard warning does not absolve the platform of liability. If this defense were accepted, they argued, victims of AI-generated false statements would be left virtually defenseless, because the false claims never appeared in any original source, meaning no third-party source could be held legally responsible for the harm.

The court further ruled that AI-generated results cannot be protected under free speech principles, because they are the product of an algorithm designed, trained, and controlled by a private company, not an individual expression of personal opinion.

As a precautionary measure to prevent similar harm from recurring, the ruling ordered Google to remove the vast majority of the defamatory statements at issue in the case, and to cover 80% of the plaintiffs’ legal costs from the proceedings.

A Google spokesperson told Ars Technica the company is likely to appeal the decision. “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,” the company said in a statement. “We’re carefully reviewing this decision, which is not yet final.”

The German court’s ruling is expected to have far-reaching global repercussions for the entire AI industry. Firms including OpenAI, Anthropic, and Perplexity AI already include similar disclaimers, warning users that their AI-generated responses may contain errors or misleading content, and recommend verifying information before use, just like Google. These disclaimers are almost always included in the terms of service users agree to when creating a platform account.

But this ruling establishes that such standard warnings are not sufficient to exempt AI developers from legal liability. The decision holds that when an AI generates new statements that do not appear directly in its original source material, the company that designs, trains, operates, and manages the system must assume full legal liability for any damages caused by those statements.

This story originally appeared on WIRED en Español and has been adapted from the original Spanish report.