Updated
Updated · ZDNet · Aug 3
Claude Link Exposure Spurs 4-Chatbot Privacy Guide as Anthropic Fixes Search Indexing
Updated
Updated · ZDNet · Aug 3

Claude Link Exposure Spurs 4-Chatbot Privacy Guide as Anthropic Fixes Search Indexing

3 articles · Updated · ZDNet · Aug 3

Summary

  • Conversations shared from Claude were found reachable through Google dork searches, prompting a fresh guide on locking down privacy across ChatGPT, Gemini, Copilot and Claude.
  • Anthropic said it fixed the indexing issue, but the report argues user settings still matter because public links, chat history, training permissions and memory features can keep sensitive exchanges exposed.
  • ChatGPT, Gemini and Copilot all let users limit retention or training, though Gemini’s strongest opt-out also turns off saved activity and history; Claude offers incognito chats and private sharing instead of public links.
  • Across all 4 chatbots, the main safeguards are the same: avoid public share links, delete sensitive chats, disable model-training and memory options, and review connected services or voice-recording settings.

Insights

Why do specialized platforms promise zero data retention while major tech giants quietly log your most sensitive conversations?
If human reviewers can read your deleted AI chats for years, is your private data ever truly erased?
Could the harmless questions you ask a free chatbot today become publicly searchable evidence in tomorrow's disputes?