Stories

Mon, 27 Jul 2026

Mon, 27 Jul 2026 Some people's chats with Claude AI found publicly available online

Hundreds of conversations with Anthropic’s chatbot were discovered as being publicly accessible.

* Hundreds of conversations with Anthropic's AI chatbot Claude were made publicly accessible through Google search results.
* Users who shared links to their chats had inadvertently allowed them to be indexed by search engines, making them visible to anyone.
* The conversations included personal and work information, such as resumes, corporate projects, and proprietary research.
* Some users used the chatbot for seemingly frivolous purposes, while others conducted serious discussions on topics like cloud security and healthcare.
* Anthropic has stated that users have control over whether or not to share their chats, but critics argue that the "share" option does not adequately warn users about the potential consequences of sharing a link.
* This is not an isolated incident - similar issues have occurred with OpenAI's ChatGPT and Grok, an AI chatbot within Elon Musk's social platform X.