WorkJuly 27, 2026via TechCrunch AI
PSA: Your Claude shared chats and Artifacts may have ended up on Google
Why it matters
A significant privacy and data governance issue at a major AI lab exposes the operational risks of consumer-facing AI products. This affects user trust and raises questions about how AI companies handle sensitive conversation data.
Key signals
- Claude's 'share chat' feature created publicly indexable URLs
- Shared chats and artifacts were discoverable via Google search
- Users assumed shared links were private but were actually searchable
- Issue originated from feature design flaw in URL accessibility controls
- Impacts privacy expectations for AI product users
The hook
Anthropic's Claude exposed shared chats to Google indexing — a privacy breach affecting users who thought their conversations were private.
The issue appears to have originated from Claude’s “share chat” feature, which allows users to create links that enable anyone with the assigned URL view a conversation or project.