If an AI chatbot misleads you, who is to blame?

12 days ago

A German court ruled Google is liable for AI search summaries, rejecting defenses that users can verify or should not blindly trust AI-generated information.
Internet companies have historically straddled the line between carriers (transmitting content without liability) and publishers (responsible for content they curate or create), with Section 230 shielding them from liability for others' speech.
AI summaries, like Google's AI Overview, involve rewriting content with editorial discretion, making companies more akin to publishers and thus liable for inaccuracies or harm, as seen in Air Canada's case where a court held the airline responsible for its chatbot's promises.
Liability for AI agents is crucial to ensure accuracy and accountability; allowing companies to hide behind faulty AI would create harmful incentives, undermining trust in systems like Visa's planned AI purchasing agents.
If the German ruling stands, it could force Google to improve its AI system due to high error rates, potentially making some AI applications commercially unviable but benefiting users by ensuring reliability.

Hasty Briefsbeta