A regional court in Munich delivered a landmark decision on July 22, 2025, finding Google directly liable for false statements produced by its AI Overviews feature. The case began when two Munich‑based publishers sent a cease‑and‑desist letter alleging that the AI had tied their businesses to scams, subscription traps, and other shady practices. According to the court, the AI did not merely list existing search results; it rewrote information in its own words, creating new, substantive statements that misrepresented the publishers.

The judges classified Google as a direct infringer because the AI Overviews are considered the company’s own content, not a neutral conduit for third‑party material. The court noted that the AI combined data from unrelated sources, fabricated connections that never appeared in any linked pages, and even invented claims without any factual basis. One of the overviews described a publisher as "known for dubious business practices," a characterization that the court found unfounded.

In response, Google argued that users could verify the AI’s claims by consulting the linked sources and that users should be aware that AI‑generated summaries are not infallible. The company cited a study suggesting that only about one percent of users click on source links after reading an AI overview. Nonetheless, the court rejected the defense, emphasizing that the AI’s independent language and structure place the responsibility squarely on Google.

The ruling includes a temporary injunction that prohibits Google from disseminating the false overviews about the two publishers. By treating the AI summaries as original content, the decision diverges from existing German laws that typically shield search engine operators from liability for third‑party material. Legal experts see the judgment as a potential turning point for how AI‑enhanced search services are regulated across the European Union.

Google’s Gemini‑powered AI Overviews attract roughly 2 billion user interactions each month, according to the company’s own figures. A New York Times‑cited study estimates that the feature gets factual information wrong about nine percent of the time. While the percentage may appear modest, the sheer volume of queries translates to millions of erroneous answers annually. Moreover, a separate analysis found that 56 percent of correct answers could not be corroborated by the linked sources, raising concerns about transparency and user trust.

Industry observers warn that the Munich court’s decision could prompt other jurisdictions to scrutinize AI‑generated content more closely. If courts elsewhere adopt a similar stance, tech companies may need to overhaul how they present AI summaries, possibly adding clearer source citations or limiting the scope of automated content generation.

For now, Google must comply with the injunction while it evaluates its legal options. The case underscores the growing tension between rapid AI deployment and existing legal frameworks designed for traditional search engines.

This article was written with the assistance of AI.
News Factory APP - agentic news to boost your SEO & AEO.