In a nutshell: Anthropic enabled shareable links to chat conversations that search engines indexed without noindex tags, making sensitive data publicly available.
Hundreds of shared chat conversations from Anthropic’s AI model Claude were publicly discoverable via Google, Bing and other search engines. The conversations contained job applications, business data and access credentials.
Reddit users discovered that targeted search queries using the operator site:claude.ai/share revealed at least 200 conversations across at least 25 search results pages on search engines. The affected chat conversations were those where users had activated the integrated sharing function to generate a public link. This affected multiple search engines simultaneously: Google, Bing, Brave and DuckDuckGo displayed the links in their search results.
The indexed content included sensitive corporate and personal data: job applications with names and contact information, drafts for corporate blogs with project details, access credentials for cryptocurrency wallets, and transcripts of internal research conversations from the healthcare sector. This disclosure represents a significant security risk for affected companies and individuals, particularly because the data was accessible through standard search queries for an extended period.
Anthropic emphasized in a statement that users retain full control over sharing their conversations. According to the company, the generated links are not guessable and not discoverable as long as users do not actively share them. The statement implies that Anthropic is shifting responsibility to users who activate the sharing function. Google pointed to its responsibility as a search engine and emphasized that website operators have control mechanisms such as noindex tags and crawl directives available to prevent indexing.
Following disclosure of the incident, the links were removed from search results. To prevent such cases, website operators have technical measures available: noindex meta tags in HTML source code or robots.txt files can specifically prevent search engines from indexing certain pages or directories. Anthropic could have implemented these protective measures on the /share directory by default to exclude sensitive chat conversations from indexing in the first place.
Similar incidents have occurred in the past with other AI offerings: in the previous year, comparable search engine indexing was documented at OpenAI’s ChatGPT and Grok’s X service. These recurring incidents point to a systemic problem with security configuration of public sharing functions in AI chatbot platforms.
Source: www.it-daily.net · Published 28 July 2026
Lumi AI News — AI-assisted curation pursuant to Article 50 EU AI Act. Paraphrase and classification by Lumi News Pipeline v1.7.3.