-
According to the CEO of the chatbot with artificial intelligence, certain users were able to view the titles of other users’ talks due to a ChatGPT bug. Users published screenshots of chat histories that they claimed weren’t theirs on the social media sites Reddit and Twitter. Sam Altman, the CEO of OpenAI, claimed the company is “terrible,” but the “major” blunder has been corrected.
But many users continue to worry about their privacy on the platform. Since its debut in November of last year, ChatGPT has been used by millions of individuals to write songs, code, and even draft messages. Every interaction with the chatbot is recorded and accessible later in the user’s chat history bar.
Yet, people started to see chats in their past that they claimed they hadn’t had with the chatbot as early as Monday. On Reddit, a user provided a picture of their chat history, which included chats in Mandarin and subjects like “China Socialist Growth.” The business informed Bloomberg on Tuesday that it had temporarily disabled the chatbot late on Monday to correct the problem.
Additionally, they claimed that users were unable to access the real chats. The CEO of OpenAI stated that there would soon be a “technical postmortem.” Users are worried about the issue because they believe the tool could leak their personal information. The bug appeared to suggest that OpenAI is able to view user communications.
Google and Microsoft, a significant OpenAI investor, have been vying for dominance of the rapidly expanding artificial intelligence tool market. Yet, given how quickly new products are updated and released, many people are worried that mistakes like these could be dangerous or result in unforeseen effects.
According to the company’s privacy policy, user information like prompts and responses may be used to keep the model under development. Yet, that information is only utilized after PHI has been eliminated. The error also occurs the day after Google introduced Bard, their chatbot, to certain beta testers and journalists.