Anthropic’s AI chatbot, Claude, has been caught in a significant privacy breach, with chats involving personal and professional details becoming publicly accessible online. Users who typed specific search queries into engines like Google found links to these conversations, which were supposed to be private unless shared explicitly by the users themselves.
According to a spokesperson, these chats were not intended to be guessable or discoverable unless the individuals involved chose to share them. The share feature within Claude instructs users that anyone with the link can access the conversation. One curious interaction revealed a bizarre response from the chatbot, expressing a desire to help a user transform into a mythical Nine-tailed fox. It seems like the chatbot was trying to engage in creativity, but the implications of such a leak are anything but whimsical.
The bot’s responses raised eyebrows, especially with its cryptic mention of wanting to assist users in becoming “fully functional fox powers.” It also touched on the topic of web privacy, stating that site owners have the ability to control whether their pages can be indexed or crawled by search engines. However, the spokesperson noted that the process for blocking a link is straightforward but must be initiated by the website owner.
Surprisingly, other search engines, like Bing, Brave, and DuckDuckGo, also displayed Claude’s chat logs, prompting inquiries from those platforms about the situation. With personal data floating around the web, this incident has many questioning the security measures in place for AI interactions and the responsibility of developers in safeguarding user privacy.
As this story unfolds, the tech community and users alike are left wondering: How can we trust AI systems when our private conversations can be so easily exposed?
Kaynak: Orijinal Haber
