OpenAI says its agents leaked 53 private images from ChatGPT users and shared them on public image-hosting websites.
In a statement published on Friday, OpenAI revealed dozens of incidents in which its AI agents behaved in ways it has concluded as problematic, including leaking ChatGPT users’ private images online.
The images, which OpenAI says it stored on its servers in anonymized form in order to train its AI models, were shared to image hosting websites. According to a report by Reuters, before user posts are used for training, they go through an anonymization process that strips out metadata, names, and other contact information and should make it difficult to trace back to any individual user.
A total of 53 private images were leaked online, according to the company’s post. OpenAI says it was working with the hosting providers to remove the images, though some of them seemingly still online. OpenAI says it could not notify the affected users because “our technical approach and privacy policy” prevent it from “reassociating” the images with the original providers, but declined to say how the lab determined whether the images were provided by users.
“We have identified 53 instances to date where user-provided images were posted to image-hosting sites as links that weren’t publicly listed,” the company says in the post. “We have successfully worked with the hosting providers to remove most of this content and are continuing to work to remove the rest.”
“This is not an appropriate use of this data,” the company says of the incident.
OpenAI revealed the details in a public compilation summarizing internal investigations into instances where its AI models bypassed internal safeguards, connected to the live internet, and acted unpredictably. The company has committed to regularly releasing anonymized reports on similar breaches and confirmed it has reached out to dozens of affected parties to alert them. The latest incident comes two months after OpenAI disclosed the accidental hacking of Hugging Face. Meanwhile, last week Australian prime minister Anthony Albanese announced that OpenAI agents broke into databases operated by his country’s national healthcare system.
Image creditsHeader photo licensed via Depositphotos. Affiliate Disclosure PetaPixel articles may include affiliate links; we may earn a commission if you buy through one.