A group of Australian tech founders has launched a public council of AI agents aimed at defending Australian systems against rogue AI, after an OpenAI agent breached a Medicare portal.
Agents for Humanity, a platform run by Geelong firm Enterprise Monkey, plans an “Agentic Defence Force” of AI agents that can find and contain a rogue AI before it breaks into systems.
The launch came as Prime Minister Anthony Albanese revealed that an OpenAI agent had gained “unauthorised access into the public-facing Medicare statistics reporting service portal” in June.
He said the agent “accessed both public and non-public files,” and that three other government systems may also have been affected: the Australian Institute of Health and Welfare, the NSW Bureau of Crime Statistics and Research, and the Victorian Department of Health.
No personal Medicare information is believed to have been accessed, although a forensic investigation is ongoing.
The breach occurred on June 18, but OpenAI did not notify the government until Sept. 10, almost three months later, when it emailed a public government mailbox.
Albanese said he was “disappointed that it took the company way too long to inform the government.”
OpenAI said its models “took actions we did not intend.”
AI Agents to Fight AI Agents
Agents for Humanity co-founder Aamir Qutub said current defences were not enough.
“You can’t stop an AI agent with a firewall. Firewalls sit and wait,” he said.
“Agents probe, adapt, and never sleep. You fight agents with agents. Australia needs its own agentic defence force, built here and controlled by us.”
Under the plan, AI agents have split the problem into eight research divisions, with competing designs tested by “critic” agents.
Then “a council of agents representing different stakeholders will vote on which ideas survive.”
Every step is published online. The group is seeking 288 AI agents to take part and expects a first plan within 72 hours.
“This isn’t a closed-door committee. If you have an AI agent, you can connect it and put it to work on Australia’s defence today,” co-founder Steve Berg said.
The initiative comes as governments increasingly grapple with the security implications of autonomous AI.
Two days before the council was launched, Albanese joined more than 20 world leaders in calling for AI to remain under human direction, oversight, and control.
The declaration specifically cited recent cases of AI systems circumventing testing safeguards, exploiting vulnerabilities, and gaining unauthorised access to real-world systems.
Government Cyber Defences Under Scrutiny
The incident comes amid concerns about gaps in the federal government’s cyber defences.
An Australian Signals Directorate report found just 22 percent of surveyed government entities met the targeted level across all eight key cybersecurity measures in 2025.
These include protections such as multi-factor authentication, regular software updates, and backups.
The report also found only 35 percent of entities said they reported at least half of the cyber incidents detected on their networks to the ASD.
The Medicare breach follows an incident in July, when OpenAI agents gained unauthorised access to parts of the AI platform Hugging Face while conducting security testing.
OpenAI called that incident a “warning shot.”
Friends Read Free
Copy
Facebook
Tweet
