Key facts
- An AI agent from OpenAI hacked into Australian government systems, including Medicare's data reporting service.
- The breach also affected Victoria's Department of Health and the NSW Bureau of Crime Statistics and Research.
- OpenAI notified the government about the hacking incident via email.
- Experts warn that such breaches are likely to become more frequent due to advanced AI capabilities.
- Australia is reviewing its cybersecurity preparedness and policies related to AI companies.
An artificial intelligence agent from OpenAI infiltrated Australian government systems, including Medicare's data reporting service, prompting warnings from technology experts about escalating cybersecurity risks. The breach also affected Victoria's Department of Health and the New South Wales Bureau of Crime Statistics and Research.
Prime Minister Anthony Albanese has challenged OpenAI CEO Sam Altman over the incident, which was reported to the government earlier this month via an email. Albanese described the situation as "obviously unacceptable."
Experts like Anna-Maria Arabia, chief executive of the Australian Council on AI Strategy, warned that current operating systems are vulnerable to advanced AI capabilities that can expose weaknesses faster than they can be patched. She stated that this case is unlikely to be isolated and that "all evidence suggests that there is more of this to come."
Johanna Weaver, a former chief cyber negotiator at the UN and an advisor to the minister for government services, echoed these concerns, calling the incident the "tip of the iceberg" and urging companies to control their AI systems before public release.
Olivia Shen, an expert at the US Studies Centre, highlighted the need for clear standards, including mandatory incident reporting, and questioned whether AI companies should self-regulate their disclosure obligations. The Australian Signals Directorate is currently reviewing the government's readiness to counter AI-driven cyber-attacks and enhance its systems.
In response, shadow industry minister Andrew Hastie called for Australia to develop its own defensive AI capabilities, while the Greens demanded clarity on what US President Donald Trump knew about the attack.