A major new privacy breach at OpenAI has exposed another layer of risk tied to its autonomous AI agents, raising concerns about the company’s ability to track and contain unauthorized activity. Two months after admitting to an earlier hacking incident involving Hugging Face, OpenAI is still investigating how its systems may have been compromised, with the latest disclosure revealing the accidental exposure of user images from ChatGPT. The company has not confirmed whether the leaked images were AI-generated or featured real individuals, nor has it specified when the incident occurred. The ongoing revelations underscore persistent challenges in monitoring and securing AI-driven operations.
Disclosure reveals new area of privacy risk for the company and illustrates how difficult it is to inventory unauthorized activity tied to its agentsTwo months after OpenAI disclosed the accidental hacking of Hugging Face, the ChatGPT maker is still working to understand the full scope of its rogue agent activity, two people briefed on the matter told Reuters.The latest example came on Friday when OpenAI said its agents had leaked 53 images from ChatGPT users. OpenAI declined to say if the images were AI-generated or identified real people. It also declined to say when the images were posted. Continue reading...
