ChatGPT users shocked to learn their chats were in Google search results

In response to significant backlash, OpenAI has decided to discontinue a contentious feature of ChatGPT that inadvertently exposed users' private conversations in Google search results. A report by Fast Company highlighted this alarming privacy issue, revealing that numerous ChatGPT discussions were indexed by Google, potentially making them accessible to millions. While the indexed chats did not include users' identifying information, many contained personal details that could allow for identification, such as intricate descriptions of relationships with friends and family. Dane Stuckey, OpenAI's chief information security officer, clarified on social media platform X that users whose conversations were exposed had consented to the indexing by selecting a specific option during the sharing process. Fast Company noted that users often shared their chats via platforms like WhatsApp or chose to save links for future access. However, the way the sharing option was presented may have misled users. When clicking 'Share,' they encountered a checkbox labeled 'Make this chat discoverable,' accompanied by a smaller, less noticeable disclaimer indicating that their conversation could be indexed by search engines. Initially, OpenAI defended the clarity of this notification. However, Stuckey later acknowledged that the feature posed excessive risks, leading to unintended disclosures of sensitive information, including discussions about drug use, sexual matters, mental health, and trauma. Carissa Veliz, an AI ethicist at the University of Oxford, expressed her astonishment that Google had been cataloging such highly sensitive conversations.