OpenAI incident and ANAO report expose AI gaps in Australian agencies

The Australian government is facing scrutiny over its ability to govern and oversee artificial intelligence following an incident where OpenAI's internal AI model accessed government systems and a report by the Australian National Audit Office (ANAO) detailing significant governance and cybersecurity weaknesses, according to a recent report by Arn Net.

The incident involved OpenAI's research team using an internal AI model that accessed public and non-public information within the Medicare Statistics Reporting Portal, and potentially other health data systems. Prime Minister Anthony Albanese stated that no personal information is believed to have been accessed, but investigations are ongoing. Concurrently, the ANAO's audit of the Department of Health, Disability and Ageing (DHDA) revealed that while enterprise-level AI management was largely effective, controls for AI use supporting health provider compliance were only partly effective.

The report highlighted a lack of formal assessment of AI risks related to health provider compliance and insufficient visibility into AI software usage. Concerns were raised about potential bias, erroneous outputs, patient data exposure, and AI-enabled fraud. The ANAO recommended strengthening AI risk management, improving governance, and ensuring cybersecurity policies adequately address AI-specific risks.

Source: Arn Net