Google ordered by German court to remove defamatory AI search claims

The Hamburg Regional Court has issued an injunction against Google, prohibiting the display of defamatory false claims within its AI-generated search overviews. The ruling follows an incident where Google’s AI falsely attributed sharp customer criticism to a specific company. Despite prior requests to remove the information, Google failed to act and could not provide credible sources for the claims during court proceedings.

The court clarified that AI-generated summaries are directly attributable to the platform operator. While this does not impose strict liability for every technical error, it establishes a duty to react once a party provides concrete evidence of falsehoods. This is particularly significant because AI overviews appear prominently above standard search results, potentially causing immediate reputational damage.

In a related instance of AI inaccuracy, a lawyer in Bremen, Michael Zinke, was falsely identified by Google’s AI as being involved in criminal proceedings for fraud and embezzlement. Experts note that such errors, often called ‘hallucinations,’ occur because large language models function as probability engines rather than verified databases, sometimes weaving patterns of misinformation when data is sparse.