Defamation and AI Search Overviews: When Google's AI Gets It Wrong

Google's AI Overviews now appear at the very top of search results — above traditional links — offering a synthesised summary drawn from multiple sources. When that summary contains false information about a real person or business, a new and significant category of reputational harm is created. This guide examines the emerging UK legal position.

What Are AI Overviews and Why Do They Matter for Defamation?

Google's AI Overviews synthesise content from multiple indexed sources and present a composite summary under Google's own name at the top of search results. Unlike ordinary search results — where Google is merely a directory pointing to third-party content — AI Overviews are created by Google's own technology. Where that synthesised content contains false statements of fact about an identifiable person or business, Google may be the primary publisher of those false statements, rather than merely a host of third-party content.

The Publisher vs Host Distinction

Under UK defamation law, a publisher who creates and disseminates content is treated very differently from a host who merely provides storage for third-party content. Google's traditional search results have been treated as secondary publication — hosting third-party links without editorial input. AI Overviews are different: the false statement is generated by Google's AI, even if it draws on third-party source material. The better legal argument is that Google is a primary publisher of AI Overview content, making it potentially fully liable for false statements it generates.

AI Chatbots: ChatGPT, Gemini, and Copilot

The same analysis applies to false information generated by AI chatbots. OpenAI's ChatGPT, Google's Gemini, and Microsoft's Copilot regularly generate confident false statements about real people — a phenomenon known as "hallucination." Where those statements cause serious harm to an identifiable person's reputation, the operator of the AI system may be the publisher of the defamatory statement. UK law has not yet definitively resolved this question, but the direction of travel — in the UK and internationally — is increasingly clear.

Getting False AI Content Removed

Google's feedback mechanism on individual AI Overviews allows reporting of inaccurate content. A formal legal removal request via Google's legal portal, supported by a solicitor's letter citing the Defamation Act 2013, is more effective for serious cases. UK GDPR rights of rectification and erasure for inaccurate personal data provide a parallel lever against AI operators.

Related reading: Algorithm amplification defamation | Removing defamatory content | Online Safety Act 2023