German courts split on Google AI liability

What happened

A Berlin court ruled in early June that Google's AI Overviews are simply a new search result format that aggregates content from other websites, not independent statements Google controls. Days earlier, a Munich court reached the opposite conclusion, holding Google directly liable for false claims made by its AI, arguing that Google's control over the AI model and algorithms makes the AI-generated text independent content.

Why it matters

The two rulings rest on fundamentally different legal grounds—Berlin focused on trademark and competition law and found no infringement, while Munich addressed false factual claims and held Google responsible because users cannot verify the information themselves and AI summaries often contain claims not found in any linked source. The contradiction means the core question of liability for generative search remains unresolved, creating uncertainty for Google and any company selling AI models with internet access.

What to watch

How appeals courts handle these questions could reshape the entire business model of AI search, not just for Google but for every company selling AI models with internet access. The practical gap between what courts say users will do (verify sources) and what they actually do (treat AI summaries as complete answers) may ultimately determine which legal framework prevails.