OpenAI unintentionally released 53 images via ChatGPT
OpenAI has acknowledged that it made a mistake by leaking 53 images through the ChatGPT platform, after inadvertently publishing links to users' personal photos on electronic hosting sites. This was due to unintentional behavior by AI agents. The company explained that the images were sent from accounts that had agreed to share their data to improve the models, and most of them have been processed and deleted by the service providers. They confirmed that the images are no longer linked to the users, and that the company is working to remove any remaining ones. CEO Sam Altman admitted that their response and detection of the incident have been slow.
Size: 120%Text Size
Notice: This Is an AI-Generated Summary
Comments (0)