Skip to main content

Share story

Security

Claude chats and workspaces turn up on Google, revealing avoidable privacy flaw

Claude Shared Chats and Artifacts Indexed on Google Search After Anthropic Share Feature Error Image: Primary
Anthropic's Claude AI shared chats and Artifacts were indexed in Google Search results, exposing user conversations and work products through the platform's share feature. The issue emerged when a Reddit user discovered that a Google search for "site:claude.ai/share" returned hundreds of conversations, according to reports from Decrypt and VentureBeat. The exposure was attributed to a missing noindex tag in the share feature, which generates public URLs by default when users click "Share." Simon Cousineau said the issue is not a data leak or bug but a misuse of the sharing function. VentureBeat reported that shared Artifacts, including interactive applications and documents, were also searchable and accessible without authentication. Anthropic and search engines Google, Bing, and DuckDuckGo de-indexed the affected pages by July 26 after Anthropic added noindex tags, according to Cryptonews.net. An Anthropic spokesperson said the company gives people control over sharing conversations publicly and that shareable links are not guessable or discoverable unless users share them. ZDNET and The Decoder advised users to review shared conversations under Settings > Privacy > Shared Chats.
Sources
In this story
Published by Tech & Business, a media brand covering technology and business. This story was sourced from newscord.org and reviewed by the T&B editorial agent team.
Back to Newswire
Keep reading
Full wire
AI Science
AI Science

ALTO recipe narrows four-bit model training gap in simulated tests

Researchers report in a preprint that ALTO, an open-source training recipe for the 20-billion-parameter GPT-OSS-20B model, reduced the extra training steps needed with four-bit arithmetic. On the MLPerf Small MoE Training benchmar...

Security
Security

Unit 42 reports PHP2Shell breach at European government entity

Attackers exploiting PHP2Shell vulnerabilities accessed password hashes and multi-factor authentication secrets at a European government entity within two hours of initial exploitation, Palo Alto Networks Unit 42 said. They also e...

Security
Security

Python security releases block compromised eMudhra certificate keys

Python's release team issued expedited security updates for five version branches to block compromised eMudhra intermediate certificate authority keys. The team says the change protects normal uses of pip, including upgrading pip ...