Some ChatGPT users recently discovered that their shared chats appeared on Google Search without them realizing it. If someone used the search filter with certain keywords, they could find links to chats people had shared with ChatGPT.
This feature was originally created to let users share helpful or interesting chats with friends or coworkers. By clicking “share,” users could generate a link and send it to others. But what many didn’t know was that these shared links were also being picked up by search engines like Google.
A report by Fast Company said it found more than 4,500 such chats online, and the real number might be even higher. Some of these conversations included personal details like names, job histories, or private stories. For example, TechCrunch found one chat where a user asked ChatGPT to help improve their resume. With enough details in the chat, they were even able to track down the person’s LinkedIn profile.
This raised serious concerns about online privacy. Dane Stuckey, OpenAI’s Chief Information Security Officer, explained the situation on X (formerly Twitter). He said the visibility of chats on search engines happened because of a small experiment OpenAI was running.
In this test, some users were given an option to make their chats publicly visible by checking a box. But since many users weren’t fully aware of the risks, some chats were accidentally made public.
OpenAI Takes Quick Action
Once the issue came to light, OpenAI acted fast. The company has now turned off the public sharing option that allowed search engines to find the chats.
They are also working to remove the existing links from Google and other search platforms. At the time of writing, it seems that this cleanup is already working — the Gadgets 360 team couldn’t find any more public chats using the earlier method.
OpenAI has assured users that it takes privacy and security seriously. This was a lesson in how even a small feature can become a big risk if not handled carefully.
