%20--%3e%3cdefs%3e%3cstyle%3e%20.st0%20{%20fill:%20%23061b40;%20}%20.st1%20{%20fill:%20%23306af1;%20}%20.st2%20{%20fill:%20%235ce5cf;%20}%20%3c/style%3e%3c/defs%3e%3cg%3e%3cpath%20class='st0'%20d='M55,10.5h9v3h-9v9h12V7.5h-12v3ZM64,19.5h-6v-3h6v3Z'/%3e%3cpolygon%20class='st0'%20points='69%2016.5%2078%2016.5%2078%2019.5%2069%2019.5%2069%2022.5%2081%2022.5%2081%2013.5%2072%2013.5%2072%2010.5%2081%2010.5%2081%207.5%2069%207.5%2069%2016.5'/%3e%3cpolygon%20class='st0'%20points='95%2010.5%2095%207.5%2083%207.5%2083%2022.5%2095%2022.5%2095%2019.5%2086%2019.5%2086%2016.5%2095%2016.5%2095%2013.5%2086%2013.5%2086%2010.5%2095%2010.5'/%3e%3cpath%20class='st0'%20d='M40,1.5v21h11.6l1.4-1.4v-7.6h0c0,0-1.4-1.5-1.4-1.5l1.4-1.4V2.9l-1.4-1.4h-11.6ZM50,19.5h-7v-6h7v6ZM50,10.5h-7v-6h7v6Z'/%3e%3c/g%3e%3cpath%20class='st1'%20d='M23.1,24L14.7,4.8l-4.9,11.2h3.8l-1.8,4H3.3L12.1,0H2C.9,0,0,.9,0,2v20c0,1.1.9,2,2,2h21.1Z'/%3e%3cpath%20class='st2'%20d='M34,0h-16.8l10.6,24h6.2c1.1,0,2-.9,2-2V2C36,.9,35.1,0,34,0ZM32.5,20h-4V4h4v16Z'/%3e%3c/svg%3e)
Recently, Anthropic's Claude AI was involved in an unsettling privacy incident: a large number of user-generated session sharing links were publicly indexed by search engines such as Google and Bing, allowing anyone to view the conversation content directly through searches. This was not a breach of the backend database, but rather a technical configuration oversight by Anthropic—failing to effectively prevent search engine crawlers from accessing the sharing links.
The issue lies in the huge gap between users' understanding of "sharing" and the reality of the internet. Many users believe that sending a sharing link to colleagues or friends is safe, but they are unaware that once the link appears on public web pages, forums, or social platforms, search engine crawlers may follow it and index the content.
After testing, some netizens found that the indexed sessions contained a large amount of sensitive information—wallet private keys, illegal consultation records, internal company tasks, and content that could identify personal identity—exposed openly in the search results.
Technical blunder leads to serious consequences: two configurations contradict each other
The root cause was a "self-defeating" technical situation. The /share/ path of Claude was set to be disallowed from being crawled in robots.txt, but the page response header also set the X-Robots-Tag: none attribute (equivalent to noindex, nofollow). This sounds like a double safety measure, but in fact, it backfires—Google's official documentation clearly states that if a page is blocked from crawling by robots.txt, search crawlers cannot read the noindex tag, resulting in crawlers failing to access the main text, but still discovering the URL through external links and including it in the search results. Users can simply click the link to view the complete session content.
Currently, Anthropic has communicated with major search engines to remove the relevant links and has revised its strategy to completely prohibit crawling. However, some cached or residual links still exist. This is not an isolated case—many sharing links of ChatGPT in 2025 were also indexed by Google, and earlier, Baidu Netdisk's public sharing also led to a large-scale exposure of user files due to similar issues. After repeated failures, a sharp question emerges: AI companies emphasize security and privacy, yet fail to handle basic technical configurations properly. Why should users trust them?