Particle.news
Download on the App Store

Munich Court Says Google Can Be Liable for False AI 'Overviews'

The ruling treats AI summaries as Google's own statements and could open the company to legal claims and wider regulatory scrutiny.

Overview

  • A Munich regional court issued a preliminary injunction on Friday requiring Google to stop repeating specific false claims in its AI Overviews and to pay about 80 percent of legal costs in the case.
  • Judges found the Overviews create new, self-contained statements because the AI summarizes results in its own words, evaluates content, and structures it in ways that went beyond the linked sources.
  • Two Munich publishers sued after the Overviews wrongly tied them to scams and subscription traps by mixing information from other companies with the plaintiffs' names.
  • Google said it disagrees with the ruling, will appeal, and argued the decision focuses on narrow errors while most AI Overviews are accurate.
  • Legal commentators say the ruling could narrow intermediary immunity by treating generative-AI outputs as operator speech and may affect U.S. rules like Section 230, raise compliance costs for AI services, and prompt closer regulator scrutiny.