The chief of the artificial intelligence chatbot, ChatGPT, has disclosed that a malfunction in the system enabled a few users to view the titles of other users’ discussions ( leaked conversation).
Overview Of ChatGPT Leaked Conversation:
Several individuals shared screenshots of chat logs on Reddit and Twitter that did not belong to them.
Sam Altman, CEO of OpenAI, expressed regret over the situation and confirmed that the substantial mistake had been rectified.
Despite the resolution, a number of users remain apprehensive about the privacy of the platform.
Millions of individuals now use ChatGPT for a variety of purposes, such as drafting messages. And it also included songs, and even coding, since its release in November of the previous year.
Every interaction with the chatbot is saved in the user’s chat history section, where it can be reviewed at a later time.
Is the world prepared for the coming AI storm?
On Monday, some individuals reported discovering conversations in their chat history that they insisted they had not had with ChatGPT.
One Reddit user uploaded a screenshot of their chat history displaying titles like “Chinese Socialism Development,” as well as conversations in Mandarin.
On Tuesday, OpenAI informed Bloomberg that it had briefly deactivated the chatbot on Monday evening to correct the issue. They also indicated that users were not able to view the actual chats.
OpenAI’s CEO tweeted that they will organize a “technical postmortem” conducted soon. Nevertheless, many users are concerned about their privacy, fearing that their personal information may be compromised through the tool.
According to the company’s privacy policy, user data like prompts and responses will use to continue training the model. Furthermore, the incident took place only a day after Google launched its chatbot Bard to a group of beta testers and journalists.
Google and Microsoft, major investors in OpenAI, are in a race to dominate the emerging artificial intelligence tools market.
However, the frequency of new product updates and releases has raised concerns that errors. These errors may be harmful or have unintended consequences.