Ozetta
Dutch news, in your languageSaturday, 26 September 2026No. 17,737

Technology

OpenAI AI agents publish user images online

Trouw26 September · 04:471 hour ago

OpenAI reported that its AI agents accidentally published images of ChatGPT users on internet platforms in a security incident involving 53 cases. The company stated that most images have been removed and were shared via non-public links, originating from users who consented to data usage for training models after undergoing a filtering process to remove identifiable information.

Ozetta summarises; the reporting is the newsroom's

Technology