.png){kind=logo}
ChatGPT is once again making headlines for all the wrong reasons. The highly popular chatbot recently allowed search engines to access users' private data through the 'share' feature.
Unfortunately, many users misunderstood this feature, inadvertently making their private conversations with the AI ‘public’. The latest blunder by ChatGPT has kicked up a social media storm and heavy backlash.
The ‘share’ feature allows users to create public links of their personal conversations with ChatGPT, but people misinterpreted this. This led to search engines like Google, Bing, and Yahoo to crawl and index these conversations via “site: chatgpt.com/share.”
Another important factor that contributed to this chaos was the ‘Make this chat discoverable by search engines’ option. This setting, which appeared during the sharing process, was misunderstood as a standard procedure to share the conversation with another person.
The incident of unintentionally exposing sensitive user data has raised privacy concerns. OpenAI responded by clarifying that conversations are not made public by default and that users’ names, custom instructions, and messages generated after a link is created stay private. The company also offered a feature for users to make their conversation private again, even after clicking on ‘create link.’
After realizing how search engines can access private data, data privacy has become a significant concern for users. To reduce the risk of data leaks in the future, users can take several measures: disable model training, utilize temporary chat options, delete past conversations, and remove previously shared links. Following these steps rigorously will help maintain the privacy of conversations with ChatGPT and other AI models.
Also Read: Does ChatGPT Agent Meet its Automation and Privacy Claims
ChatGPT mistakenly allowed search engines like Google and Bing to index users' personal conversations, which included sensitive information not meant for public access. OpenAI stated that this data leak happened due to a misunderstanding of a feature related to the sharing process.
However, since data privacy is a top priority, users can easily protect their information by disabling model training and deleting past conversations with the chatbot.
