German ruling declares Google liable for false answers in AI Overviews
Points and comments are a snapshot, not live.
German court rules Google is liable for false statements in AI Overviews, rejecting search engine immunity.
Munich Regional Court found Google directly liable for defamatory AI Overviews that falsely linked two publishers to scams. The court classified AI-generated summaries as Google's own content, not mere search results, because Google controls the algorithms and rewrites information in its own structure. The AI made claims not present in linked sources. Google's defense that users could verify sources themselves was rejected; the court noted studies show under 1 percent of users click sources. The ruling bars Google from spreading the false claims and makes it pay 80 percent of legal costs.
The court found existing search engine liability exemptions do not apply because AI Overviews generate independent statements, not just point to external content. An analysis cited in the article found Google's AI answers correctly 91 percent of the time, but at scale this means millions of wrong answers hourly. Fifty-six percent of correct answers could not be traced to linked sources.
What commenters are saying
Top comments debate whether the ruling signals the end of AI Overviews in the EU and whether the precedent extends to ChatGPT, Claude, and Perplexity. Most agree the distinction is sound: search engine immunity applies to indexing third-party content, but AI summaries are Google's own authored statements and should face defamation liability like any publisher. One commenter notes the ruling may spawn legal actions against other AI providers. Skeptics worry overregulation will push innovation away from Europe. Few dispute the core logic that making false factual claims (as opposed to opinion) should carry liability regardless of medium.