Anthropic—Claude's share feature exposed user conversations to search engines, contributing to over 100,000 indexed chats across LLM platforms
Between 2024 and 2025, Claude's 'share' feature allowed users to make conversations 'discoverable,' which resulted in these chats being indexed by search engines and archiving services. As part of a cross-platform issue affecting multiple LLM providers, over 100,000 shared chats were reportedly indexed and later scraped, exposing API keys, access tokens, personal identifiers, and sensitive business data. Users did not adequately understand that 'discoverable' meant publicly searchable and permanently archived.
Scoring Impact
| Topic | Direction | Relevance | Contribution |
|---|---|---|---|
| Consumer Protection | -against | primary | -1.00 |
| Corporate Transparency | -against | secondary | -0.50 |
| User Privacy | -against | primary | -1.00 |
| Overall incident score = | -0.715 | ||
Score = avg(topic contributions) × significance (high ×1.5) × confidence (0.57)
Evidence (1 signal)
Claude shared conversations indexed by search engines as part of cross-platform data exposure affecting 100,000+ chats
Security research revealed Claude's share feature made conversations 'discoverable' to search engines, contributing to cross-platform issue where over 100,000 LLM chats were indexed and scraped. Exposed data included API keys, access tokens, personal identifiers, and business data. Users misunderstood privacy implications of 'discoverable' setting.