Nobody needs AI to search the Internet, court says in ruling against Google

A German court has found Google liable for false statements generated by its AI Overviews feature. This landmark ruling emerged from a case where Google’s AI incorrectly linked publishers to fraudulent activities and failed to rectify the misleading information, even after receiving a cease-and-desist letter. This decision could have significant implications for all AI search engines and chatbots.

Google argued that users understand AI outputs require verification, but the court disagreed. It distinguished AI Overviews from traditional search engines, noting that AI Overviews make "independent, new, and substantive statements" based on their own interpretation of web links. The court emphasized that only Google can correct the underlying algorithm and its outputs.

This ruling establishes that AI firms cannot attribute false claims to fake sources without assuming liability, as the court stated that "nobody needs AI to search the Internet." The court also dismissed Google's argument that users shouldn't blindly trust AI overviews, highlighting that such distrust would diminish the AI tool's utility significantly.

Numerous reports indicate the inaccuracy of AI Overviews. A Pew survey revealed that most users do not click on source links, and a New York Times analysis found AI Overviews to be inaccurate about 9% of the time, with inaccurate source links appearing in 56% of cases. These findings suggest a high volume of daily misinformation.

This judicial precedent could lead to a wave of lawsuits against major AI developers if other courts adopt similar stances on liability for defamatory AI-generated content. It remains uncertain whether Google will appeal the decision or prioritize addressing requests to correct false statements more promptly.