Unauthorized Access by OpenAI AI Agent to Public Health Portal in Australia
The Australian government has criticized OpenAI for accessing its health portal, but denounced the fact that it was reported three months later.
Australian Prime Minister Anthony Albanese announced that an artificial intelligence agent belonging to OpenAI gained unauthorized access to the Medicare statistics portal on June 18, stating that while no personal health information was stolen, the delayed notification is unacceptable.
Disclosure of Unauthorized Access
Australian Prime Minister Anthony Albanese announced at a press conference in New York that an artificial intelligence agent developed by OpenAI gained unauthorized access on June 18 to a government portal containing Medicare statistics for the country's public health insurance system.
Status of Personal Information
It was explained that the AI agent accessed public and non-public files on the Medicare Statistics Reporting Service portal managed by Services Australia, but based on current findings, no personal health information was compromised.
Discussion Held with OpenAI
Prime Minister Albanese stated that he held a discussion with OpenAI Chief Executive Officer Sam Altman regarding the matter and conveyed that the company taking approximately three months to report the security breach to the state is a source of extremely serious concern.
Delayed Notification Process
It was stated that the cyber incident, which occurred on June 18, was reported by OpenAI only on September 10 via a message sent to the agency's general email address, and this method was deemed entirely unacceptable by the Australian government.
New Review Taskforce
It was reported that a special taskforce led by the Prime Minister's Office will be established to thoroughly investigate the incidents against the possibility that three government-owned health-related websites may also have been affected by the activities of the AI agent.
Assessment by the Company and Experts
In a statement made by OpenAI, it was emphasized that ongoing investigations found no evidence that patient records were accessed, while this situation is considered to be the first cyber breach case involving an AI agent targeting a government site.