German Court Rules Google Liable for AI Overview Errors

What did the German court rule against Google?

The Regional Court of Munich issued a temporary injunction against Google on June 10, 2026, barring it from repeating false statements about two Munich-based publishers. The court ruled that Google is directly liable for false claims its AI Overviews make. It treated those summaries as Google's own speech, not ordinary search results.

Google's AI Overviews had wrongly tied the two publishers to scams, subscription traps, and "dubious business practices." The AI mixed up those publishers with genuinely shady companies and invented connections that appeared in none of the linked sources. The publishers sent a cease-and-desist letter. Google did not respond adequately, the court found.

Why doesn't existing search-engine liability law protect Google?

German courts had previously granted traditional search engines limited liability. The reasoning: they merely point to third-party pages. The Munich court drew a sharp line between that and what AI Overviews do.

AI Overviews generate "independent, new, and substantive statements," the court said. They rewrite and judge results in Google's own words and structure. Because of that, the court classified Google as a direct infringer — not a passive conduit. The court called the false claims "the defendant's own statements."

The court also ruled that Google cannot fall back on Digital Services Act host-provider protections, according to The Next Web.

What arguments did Google make, and why did they fail?

Google argued that users generally understand AI outputs aren't always accurate and can verify claims by checking the linked sources. The court rejected this.

You might also like

Meta Donates Ray-Ban AI Glasses to 130,000 Blind Veterans

OpenAI Hit With Multi-State AG Subpoena

Meta Cuts 8,000 Jobs as Zuckerberg Warns AI Success Isn't Guaranteed

Sam Altman Cancels Korea Trip, Tools for Humanity Cuts Jobs

The ability to disprove a statement through further research does not exempt whoever published it, the court said. It drew a parallel to press law: a misleading teaser is actionable even if no one reads the full article. That reasoning is backed by data — studies have found barely 1% of users click a source link from an AI Overview.

The court also weakened free-speech protections for AI output. An AI's opinion, it wrote, is "not the expression of an acquired conviction" but "the result of an algorithm" — largely an expression of Google's business interests.

Who pays, and what are the cost details?

Google was ordered to cover 80% of the legal costs. The two plaintiff publishers each pay 10%.

The ruling is a preliminary injunction from a regional court, not a final judgment. It is not binding precedent under Germany's civil-law system. Google can appeal.

How big is the accuracy problem at Google's scale?

Here's what we know so far from the sources: an analysis by AI startup Oumi for the New York Times found Google's AI Overviews — running on the Gemini 3 model — answered correctly 91% of the time. That sounds high. But at Google's volume, it still means millions of wrong answers every hour.

The Next Web also noted that more than half of even the correct answers were not supported by the sources cited in the Overview.

Which other AI products could this ruling affect?

The court said its reasoning could have international reach. The same logic, if it survives appeal, would apply to every AI answer engine. Ars Technica reported that the ruling could spell trouble for the broader AI search industry — including ChatGPT, Claude, and Perplexity.

The ruling lands as European pressure on Google intensifies. Google already faces a major EU fine and orders to open Android to AI rivals under the bloc's new rules.

For builders and developers working on AI search products, the liability question is now concrete, not hypothetical. The court's reasoning targets any system that generates "independent, new, and substantive statements" — which describes most AI answer engines on the market today.

The ruling also arrives as companies like Anthropic and others are scaling AI products that surface direct answers to user queries. The legal exposure identified in Munich applies wherever an AI system makes affirmative factual claims rather than linking to sources.

Even industrial AI applications — like those being built by companies such as Bezos's Prometheus — could eventually face similar questions about who owns the output of an AI system.

Key facts at a glance

Google has not commented publicly on the ruling. The next concrete step is Google's potential appeal of the preliminary injunction.