German ruling: Google liable for false answers in AI Overviews

What happened

On June 10, 2026 a German court ruled that generated answers in Google AI Overviews represent Google's 'own words' for the purposes of defamation law. Google therefore cannot claim 'safe harbor' as it does for hyperlinks.

Context and impact

The first landmark European ruling of its kind. If upheld on appeal, it opens Pandora's box: every AI hallucination about a real person becomes a potential lawsuit. Google is lobbying European regulators to exclude AI Overviews from this strict interpretation — meanwhile the EU AI Act is in its implementation phase.

Details

• Plaintiff: individual about whom AI Overviews stated false facts
• Court's argument: the output is editorial content, not 'third party'
• Consequences: defamation, right to be forgotten, and possibly Article 17 reading for AI