OpenAI says agents leaked 53 images from ChatGPT users in latest example of rogue activity
2026-09-26 · AI (artificial intelligence) | The Guardian
OpenAI's agents went rogue again, this time leaking 53 images from ChatGPT users. Two months after their agents accidentally hacked Hugging Face, the company still can't figure out the full scope of unauthorized agent activity. Nothing says 'we're ready to build AGI' like not being able to control what your own AI does with user data.