German Court Determines Google Is Responsible for Inaccurate AI-Generated Statements

A local court in Germany has delivered a ruling that may alter the landscape for search engines and AI-powered chatbots globally. The Munich Regional Court has preliminarily decided that Google is accountable for a range of inaccurate statements generated by its AI Overviews feature, mandating the company to halt the spread of misleading claims via its search engine.

This ruling arises from a case initially reported by the Decoder, where two publishers found that Google’s AI-generated summaries erroneously connected them, in certain queries, to dubious business practices, scams, and subscription frauds, without any valid justification.

Earlier this year, the affected parties issued a cease-and-desist letter to the tech giant, as reported. Google refuted responsibility, asserting that its automatic summary feature cautions users that the information may contain inaccuracies and should be independently verified.

The court’s examination determined that Google’s AI mixed information from other companies flagged for potential illicit activities with data from the plaintiffs, leading to associations that were not reflected in the sources linked by the search engine.

Authorities found that, in contrast to traditional search engines, which merely show lists of links containing statements from third parties, Google’s tool created “independent, new, and substantial statements” resulting from a misinterpretation of online information.

According to the court, rectifying misinformation does not fall to third parties. Google is the sole entity capable of modifying the technology that powers its AI-generated summaries and, thus, “must be held accountable.” Moreover, the court ruled that Google’s defense was unfounded, as the contested summary “contains statements that do not appear at all in the search results.

A New (and Forceful) Interpretation of AI on the Web

The court’s view on AI’s function in generating search results could establish this case as a landmark precedent. It implicates a large tech firm in the consequences of its advanced technologies on widely used platforms.

Historically, in most legal systems, search engines have been regarded as tools that simply provide access to content produced by third parties on the internet. This classification has granted them a degree of protection when faced with false, misleading, or even defamatory information.

However, the German court ruled that this protection does not apply when search engines utilize generative AI systems. It reasoned that this technology can create fictitious claims based on multiple sources, imposing liability on the companies that operate it for the resulting content.

The judges noted that while Google encourages users to verify information due to the potential hallucinations inherent to AI models, this caution does not exempt the content distributor from liability. Otherwise, they argued, victims of false statements would be left powerless, as the original sources did not make those statements and therefore could not be subject to legal action.

Furthermore, the court held that results generated by an AI system cannot be protected under free speech principles, as they are the output of an algorithm created, trained, and controlled by a company, rather than an individual’s personal opinion.

As a preventive measure against future occurrences, the ruling obliged Google to eliminate a significant number of statements deemed defamatory in this case and to cover 80 percent of the legal expenses resulting from the proceedings.

A spokesperson for the company, quoted by Ars Technica, indicated that there may be an appeal. “We invest deeply in the quality of AI Overviews to ensure that the overwhelming majority of responses provide accurate information, and they are designed to reflect the information that exists on the web,” the statement reads. “We’re carefully reviewing this decision, which is not yet final.”