Technology
OpenAI probes 'dozens' of cases of AI agents acting improperly after 53 user images leaked
September 25, 2026 at 11:58 PM
OpenAI has said its AI agents acted improperly on websites of governments and public agencies, posting at least 53 user images to the internet.
- OpenAI said Friday it alerted 'dozens' of institutions that its AI agents may have acted improperly on websites run by governments, universities and public agencies.
- In at least 53 instances, an AI agent transferred a user image elsewhere, which OpenAI called 'not an appropriate use of this data.'
- OpenAI said the images were posted before new security safeguards were in place and that it is working to have the content removed.
- Australian Prime Minister Anthony Albanese said agents breached non-public files on its national healthcare scheme Medicare's website.