Google faces increased legal liability in Germany as courts distinguish between traditional search indexing and AI-generated summaries. Under German law, AI-generated responses are increasingly viewed as original content created by the search engine, making the operator liable for defamatory “hallucinations” that violate corporate personality rights (Unternehmenspersönlichkeitsrecht).
Why are AI summaries treated differently than search links?
Traditional search engines act as intermediaries, providing links to third-party content. Under this model, the operator generally isn’t responsible for the content of those external pages unless they’re notified of a violation and fail to act. However, AI-generated summaries, such as those in Google’s AI Overviews, don’t just link to information—they synthesize it into new, original text.
According to legal principles applied in German courts, this synthesis transforms the search engine from a passive host into a content creator. Because the AI selects, evaluates, and reformulates data into a cohesive statement, the resulting text is attributed to the operator. This shift means the operator can’t hide behind the “mere transmitter” defense when an AI generates a false or damaging claim about a business.
How does German law handle AI defamation?
Companies in Germany are protected by Unternehmenspersönlichkeitsrecht (corporate personality rights), which is derived from Article 2, Paragraph 1 and Article 19, Paragraph 3 of the Basic Law (Grundgesetz). These rights protect a business from objectively false statements that could damage its reputation in the marketplace.
When an AI produces a “hallucination”—a confident but false statement—that labels a company as fraudulent or unserious, it constitutes a violation of these rights. If the statement is objectively incorrect and damages the company’s commercial standing, the court can issue an injunction (Unterlassungsanspruch) under Section 823 of the German Civil Code (BGB).
What is the difference between Störerhaftung and direct liability?
For years, digital platforms relied on Störerhaftung (interferer liability). This legal concept suggests that a platform isn’t the primary wrongdoer but may be liable if it contributes to a legal violation by failing to implement proper safeguards.
AI summaries change this dynamic. Because the AI creates the specific phrasing and the “opinion” presented to the user, the liability is more direct. The operator isn’t just an “interferer” who allowed a bad link to exist; they’re the author of the defamatory text. This makes it significantly easier for plaintiffs to secure injunctions and damages, as the operator’s “fault” is baked into the deployment of the AI model itself.
How does this compare to existing AI legal precedents?
| Feature | Traditional Search Results | AI-Generated Summaries |
|---|---|---|
| Role of Operator | Index/Link Provider | Content Creator/Author |
| Liability Basis | Notice-and-Takedown (Host) | Direct Liability (Publisher) |
| Legal Shield | Intermediary Protections | Limited; subject to personality rights |
| Primary Risk | Linking to illegal content | Algorithmic hallucinations |
What happens next for AI search engines in the EU?
The legal pressure on Google and other AI providers will likely intensify with the full implementation of the EU AI Act. The Act introduces strict transparency requirements and risk management obligations for “high-risk” AI systems. While the Act focuses on safety and ethics, it reinforces the expectation that AI providers must ensure the accuracy and reliability of their outputs.
Businesses are now more likely to monitor AI Overviews for “brand hallucinations.” As courts continue to rule that AI-generated text is a proprietary statement of the operator, search engines will have to implement more rigorous fact-checking layers or risk a wave of defamation lawsuits across the European Union.