The American company “Open AI”, which developed the “ChatGBT” artificial intelligence system, acknowledged the existence of a vulnerability in the system which allowed some users to see the addresses of other users’ conversations, but confirmed that worked to fix it.
And “Chat GBT” is a technology that was launched for free use last November, and it is a robot or an AI program in able to lead conversations and answer the questions posed in detailed and accurate manner, and can write songs and complete press topics depending on what the user provides From the rumors, quoted from the newspaper “Al-Sharq Al-Awsat”.
According to the British BBC, a number of users of social mediaTwitter, and Reddit, shared images of chat logs that they said didn’t belong to them.
OpenAI CEO Sam Altman described the vulnerability as “horrible” but confirmed it was now fixed.
However, many users remain concerned about this technology invading their privacy.
Every conversation made by ChatGBT is stored in the chat history of the user with whom he had the conversation, where it can be revisited in following.
But earlier on Monday, users noticed conversations in their history that they said they hadn’t had with the chatbot.
A user on Reddit shared a photo of his chat history, including titles like “The Development of Chinese Socialism” and conversations in Mandarin (the official language of China).
On Tuesday, OpenAI told Bloomberg that it briefly disabled the chatbot to fix the bug.
The company also confirmed that users were not in able to access actual chats, but only their addresses.