Faced with increasing play, OpenAi deleted a controversial chatgpt function, so that some users unintentionally show their private – and very personal – chats in search results.
Quick Company exposed the privacy problem on Wednesday and reported that thousands of chatgpt conversations were found in Google search results and probably only a sample of chats represented “visible for millions”. Although the indexation did not include identification of information about chatgpt users, some of their chats shared personal data – such as very specific descriptions of interpersonal relationships with friends and family members – may be possible to identify them, Fast Company found.
OpenAI's Chief Information Security Officer, Dane Stuckey, explained on X that all users whose chats were exposed have opted for indexing their chats by clicking on a box after she chose to share a chat.
Fast Company noted that users often share chats on WhatsApp or select the option to make a link to visit the chat later. But as Fast Company has explained, users may be misled in sharing chats because of how the text is made:
“When users clicked on 'Share', they were given an option to check a box that 'discovers this chat'. Under that, in smaller, lighter text was a warning explaining that the chat could then appear in the results of search engines.”
Credit: Chatgpt Share Box via Dane Stuckey on X
Initially, OpenAi defended the labeling as 'sufficiently clear', Fast Company reported Thursday. But Stuckey confirmed that “ultimately”, the AI company decided that the function “has introduced too many possibilities for people to accidentally share things that they did not intend.” According to Fast Company, it included chats about their drug use, sex lives, mental health and traumatic experiences.
Carissa Veliz, an AI ethicus at the University of Oxford, told Fast Company that she was “shocked” that Google “logging in” these extremely sensitive conversations.
OpenAi promises to remove Google search results
Stuckey called the position a “short -lived experiment” that opened OpenAi “to help people discover useful conversations.” He confirmed that the decision to remove the function also included an attempt to “remove indexed content from the relevant search engine until Friday morning”.
