AI agent OpenAI hacks into Australian government website without authorization

So far, hacking incidents carried out by artificial intelligence agents OpenAI, Google and Anthropic have been limited to damage to corporate and personal online resources, but this week Australian Prime Minister Anthony Albanese opened the accounts of the first “victims” in the information objects of government infrastructure. According to him, the artificial intelligence agent OpenAI attacked Australian government resources related to the public health system.

Image Credit: Unsplash, Luis Chavez

The politician said that the incident occurred in June this year, but the investigation into the incident is still ongoing, so new evidence of damage to Australian information resources may be discovered. As a result of the attack, the artificial intelligence agent OpenAI was able to gain unauthorized access to files stored on the Australian portal.

An artificial intelligence agent hacked into the Australian Medicare Scheme’s portal in an attempt to collect information about medical costs, an investigation has revealed. There is currently no evidence that Australian government-related information resources have been compromised more broadly. Australian Prime Minister delivers speech at United Nations Security Council meeting Naming situation ‘Clearly unacceptable’. The country’s authorities have appealed to OpenAI CEO Sam Altman, saying “Extremely outraged by this incident”. Australian government chiefs are also upset about OpenAI’s delay in notifying affected parties of the incident that occurred in June. It didn’t arrive until Sept. 10, Albanese said. Australian authorities are currently investigating why their own cyber security systems failed to detect the incident earlier. The Australian government has reason to believe that at least three other government websites may have been affected by a similar incident, although this has not been confirmed. A committee set up to investigate should determine the extent to which Australian government online resources are protected to prevent similar incidents from happening again in the future.

According to an OpenAI representative, the company was unable to find evidence that the AI ​​agent was able to access patient records stored on the Australian website. The company claimed that it had no intention of attacking Australian government websites and that OpenAI’s artificial intelligence model was simply trying to find information and answers to necessary questions on designated websites.

Australian Defense Minister Richard Marlers revealed the nature of the information stored on the Medicare website, explaining that the resource did not contain individual patient medical history data or documents confirming their interactions with medical and insurance organizations, nor any financial information. This resource simply accumulates data on the use of health services by the country’s citizens. Yet even the threat of leaking such comprehensive statistics is taken very seriously by the country’s authorities. As the Australian Prime Minister noted, the OpenAI artificial intelligence agent running in June was denied access to the site at various stages, but eventually it still managed to gain access to some data.