OpenAI says agents leaked 53 images from ChatGPT users in latest example of rogue activity
TL;DR
Disclosure reveals new area of privacy risk for the company and illustrates how difficult it is to inventory unauthorized activity tied to its agents Two 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.
Nauti's Take
The upside is that OpenAI now discloses these incidents, which creates an opportunity for real standards in agent monitoring. The risk is concrete: images from ChatGPT ended up online, and OpenAI cannot yet say whether real people are affected.
Teams running agents with access to user data should tighten logging and permissions now, before they face an incident of their own.
Sources