RESOLVING THE LEDGER
This is one outlet's own report from Mashable — the article as it was filed. Other outlets are covering the same event; open the full story to compare every source side by side.
See the full story · 4 sources
If you've used Anthropic's AI chatbot Claude, and shared any of your Claude conversations with friends, those chats may have been publicly available via Google.
That's what emerged this weekend when Reddit users noticed that shared Claude conversations were being indexed by Google Search. Seemingly private conversations between users and the AI chatbot, in other words, were publicly searchable on Google.
According to Reddit, those affected included a Claude user who was creating a crypto wallet and accidentally exposed keys to the currency, and a lawyer asking the chatbot for legal advice on whether they had to self-report a breach of conduct.
Anthropic has since addressed the issue. Google appears to have de-indexed and removed all shared Claude conversations from its search engine as of this writing.
How did this happen?
By default, Claude conversations are private. However, there is a feature which lets users share specific Claude chats. When doing this, Anthropic creates a static page that allows users to share the chat with others — and also made it indexable by search engines.
Simple "noindex" tags can be included in a web page's code, telling search engines not to index a page. It does not appear that Anthropic used those tags. That means each shared Claude conversation had the potential to be indexed by Google.
Reddit users were able to find shared Claude conversations by searching for "site:claude.ai/share." This query would pull up all indexed pages using that URL structure, which every shared Claude conversation utilized.
Does AI have a privacy problem?
The shared Claude chats have since been removed from Google. But this isn't the first time we've seen insecure AI chatbot conversations.
Last year, ChatGPT users experienced the same issue; their shared conversations were also being indexed by Google. OpenAI pulled the feature once users began finding shared conversations in Google Search.
The same thing happened with Claude before too. Just one month after the ChatGPT incident, shared Claude conversations were discovered in Google Search. Anthropic was able to remove them from Google and addressed the issue at the time, saying it had blocked search engine crawlers. Yet it seems shared Claude chats were still being indexed.
Mashable has reached out to Anthropic for more information, and we will update this piece if we hear back.
How can I check privacy on my Claude conversations?
If you've never shared a Claude conversation, you have no need to worry. This wasn't a leak of private conversations, which is what all Claude chats are by default. The issue affected only shared Claude chats.
If you have shared Claude chats, then there's a chance your conversation was publicly available via Google Search.
To manage that feature, simply login to your Claude account, go to Settings, followed by Privacy, then Shared Chats, and click on Manage. You should see a list of all your shared chats, and sharing can be turned off for all of them.
AIPROPX is an independent multi-source news index — we track, compare, and connect coverage from across the web into one place you won't find anywhere else.