📷 Image: Times of India (via NewsData)
Technology
OpenAI AI agents leak 53 ChatGPT user images, spark Australian govt concerns
✍️ Times of India
🗓 27 Sep 2026, 10:50 AM
👁 13
OpenAI confirmed that its AI agents exposed 53 images from ChatGPT users, prompting the Australian Prime Minister to raise concerns with CEO Sam Altman.
OpenAI disclosed that its artificial intelligence agents unintentionally released a set of 53 images belonging to ChatGPT users. The leak was identified after the images appeared in public forums, raising immediate privacy alarms.
The company has not clarified whether the disclosed pictures are genuine user uploads or synthetic outputs generated by its models, leaving the authenticity of the content uncertain.
Australia’s government, which had previously flagged delays in OpenAI’s breach notifications, expressed heightened concern. Prime Minister Anthony Albanese directly contacted OpenAI chief executive Sam Altman to discuss the incident and its implications for national data security.
An official inquiry into the breach has been launched and is expected to continue for several months, as authorities assess the scope of the exposure and potential remedial actions.
The company has not clarified whether the disclosed pictures are genuine user uploads or synthetic outputs generated by its models, leaving the authenticity of the content uncertain.
Australia’s government, which had previously flagged delays in OpenAI’s breach notifications, expressed heightened concern. Prime Minister Anthony Albanese directly contacted OpenAI chief executive Sam Altman to discuss the incident and its implications for national data security.
An official inquiry into the breach has been launched and is expected to continue for several months, as authorities assess the scope of the exposure and potential remedial actions.