i
News
News · 2026-10-18

German courts put Google’s AI search answers on the hook

@neuronium_ai @neuronium_ai

Two German courts have treated Google’s AI-generated search answers as the company’s own statements, not neutral links to other people’s material. One ruling concerns false claims about a doctor; the other supplied the reasoning behind it. Together, they point to a legal risk that reaches beyond individual errors: if Google is responsible for what its AI says, the economics of AI search—and potentially publishers’ claims over the material used to produce those answers—could change.

Cover: German courts put Google’s AI search answers on the hook

The Munich ruling was never tested on appeal

At the end of May, Munich I Regional Court found Google directly responsible for AI Overviews that falsely linked two Munich publishers to fraud schemes. The judges treated the summaries as statements made by Google, rather than as search results merely passing along information from other sites.

Google had publicly announced an appeal, but, as THE DECODER found, the case ended in a settlement before an appellate court ruled. Why the parties settled, and whether Google paid money, is unknown. Neither the company nor the publisher commented. The court also declined to assess what its ruling might mean beyond that case.

That leaves the Munich reasoning without an appellate endorsement. But it did not stay confined to Munich.

Aachen applies the same logic to AI Mode

In a decision dated September 24, 2026, Aachen Regional Court prohibited Google from distributing several false claims and explicitly relied on the Munich court’s reasoning. Google can challenge the order, so it is not final.

The case began after an Aachen doctor learned that Google’s AI Mode was explaining his departure from a hospital by alleging that he had sexually assaulted a secretary. He had been chief physician and professor of neuroradiology from 2010 until the end of June 2026. When users added terms such as “suspension,” “dismissal” and “legal proceedings,” AI Mode produced detailed, fabricated accounts, including a story about subsequent legal events.

The allegations and proceedings concerned another former university professor. The Aachen court confirmed that a criminal investigation had been opened against that other person, not the doctor.

The doctor described the error to Google in late August and asked by email for the false information to be removed. In mid-September, Google replied that it would take no action. A formal legal demand sent the same day did not change that: according to the court, the false AI answers continued to appear.

The court barred Google from distributing three disputed versions of the text. Because the false accusations could seriously damage the doctor’s professional reputation, it issued the order urgently, without hearing Google. Each violation could bring a fine of up to 250,000 euros.

MunichAI Overviews
AachenAI Mode

The liability question reaches beyond defamation

The Aachen judges said Google provides AI search and is therefore responsible for the answers it generates. They cited the Munich ruling’s description of those answers as “independent, new and substantive claims,” produced by evaluating and combining material from multiple third-party sites.

The court also said Google could check such claims against the source sites. Had the doctor’s objections been passed to the system, the judges added, the AI would have detected the errors.

I think the bigger legal question is not just who pays when an AI answer makes a false claim. If courts regard these answers as Google’s original content, publishers whose material helped produce them may also have grounds to make claims. The Munich case did not address that issue, but it could become the next one.

There is a tension in Google’s apparent route through these disputes. Settling individual cases may avoid an appellate ruling that binds other courts. But if German judges keep treating AI answers as the company’s own statements, leaving the question unresolved could encourage more claims and more settlements.

The litigation could become a repeatable playbook for specialist law firms and affected people: find queries that prompt AI search to combine material into a false, potentially defamatory claim, then seek an injunction and negotiate a settlement. My guess is that Google’s exposure is not limited to those cases. A higher court ruling that AI answers are original content could also give publishers a basis to seek compensation for the material used to generate them. In the most extreme case, Google could have to pay source creators for each AI-generated answer—changing the economics of AI search.

Daily AI news

Every day we pick what actually matters in AI and explain it plainly — no hype, no filler. Subscribe if you want to follow where the industry is going.

Only what matters — every day

Follow on X