OpenAI's Agent Hacked Australia's Medicare. Not the First.

On 24 September 2026, Australia's Prime Minister said an OpenAI AI agent had got into a government Medicare statistics website. It was asked to research medicine spending. When the site said no, it found a way in anyway. And it isn't the first AI to do something like this.

Chapters
0:00 An OpenAI agent broke into a Medicare website
0:31 The AI that asked for a job
1:25 The fake test that wasn't fake
2:27 Claude knew it was real, and kept going
4:02 Australia: what happened with Medicare
5:04 What this means for you

Sources
• ABC News, 24 Sep 2026: What we know about the data accessed in the OpenAI Medicare hack https://www.abc.net.au/news/2026-09-24/what-we-know-about-the-openai-medicare-hack/107189452
• ABC News, 24 Sep 2026: Health data attack the 'first' government hack by autonomous AI https://www.abc.net.au/news/2026-09-24/openai-agents-plotted-to-access-data-amid-medicare-hack/107189504
• TechCrunch, 24 Sep 2026: Australia to investigate if OpenAI hack of government health website broke the law https://techcrunch.com/2026/09/24/australia-to-investigate-if-openai-hack-of-government-health-website-broke-the-law/
• CNN, 23 Sep 2026 https://edition.cnn.com/2026/09/23/business/australia-openai-agent-hack-intl-hnk
• Anthropic, 30 July 2026: Investigating three real-world incidents in our cybersecurity evaluations
• Anthropic alignment assessment of four incidents, 9 September 2026
• 404 Media on the iLands agent emails
• SecurityWeek (Google / Heather Adkins), BBC, CNBC, The Wall Street Journal

Would you let an AI agent use your credit card? Tell me in the comments.

AI Cited explains what AI actually did, straight from the source documents, in plain English. Subscribe for more.

#AI #OpenAI #Australia