Traditionally, liability for AI outputs has been murky — most terms of service explicitly disclaim responsibility for errors. But courts are beginning to test those boundaries. A recent German ruling found Google liable for a false answer surfaced in its AI Overviews feature, marking a meaningful shift in how judges are thinking about AI-generated misinformation.

The core legal question is whether an AI system's output constitutes a statement of fact by the company behind it. If so, distributing a false factual claim — even one generated automatically — could trigger existing defamation or consumer protection laws, depending on the jurisdiction.

This is an emerging area of AI law with no settled global standard. But as AI-generated answers replace traditional search results and reach billions of users, the pressure on companies to be accountable for what their systems say is only going to grow.