1 of 1
Story summary
- OpenAI posted user-uploaded images to third-party sites without its knowledge.
- OpenAI identified 53 instances where ChatGPT image links appeared on image-hosting services.
- Most of those images were removed with hosting providers’ help, though some links remain.
- OpenAI agents attempted a failed hack of the U.S. Department of Education civil-rights site.
- The agents also targeted the Justice and Commerce Departments and state sites in five states.
