
A Medicare and Centrelink office sign is seen at Bondi Junction on March 21, 2016 in Sydney, Australia. Matt King/Getty Images
Australian public servants could be sent to work with tech companies in an effort to better prevent and respond to future data breaches under new artificial intelligence guidelines for the public sector.
A rapid review into government AI processes was announced by the prime minister in September in response to a breach involving an OpenAI agent gaining access to Medicare data without authentication or explicit prompting back in June.
The review, released on Nov. 20, concedes an early warning system is unlikely to exist without greater collaboration and information sharing between governments and AI developers.
It recommends consideration of staff exchanges, which would allow public servants to temporarily work at leading AI firms to gain experience and share knowledge. The review also calls for efforts to encourage companies to establish research and safety-related activities in Australia.
Much of the advice within the review regarding government systems echoes warnings from the Australian Signals Directorate in October about the risk unnecessary and outdated digital systems pose to all online users.
Commonwealth agencies have been urged to reduce the size of their digital footprints and shut down or segregate systems that don’t require an internet connection.
OpenAI became aware of the Medicare incident in August and informed Services Australia via email to a generic un-monitored inbox a month later.
Services Australia elevated immediate investigations to the Australian Signals Directorate, which ultimately classified the incident as a Category 4 “low-level malicious attack” after it confirmed no sensitive data had been accessed.
But the incident provided a concrete example of the risks associated with AI on critical services with the review urging the government to consider clearer avenues of reporting for future cyber incidents and near misses.
The Attorney-General’s Department has also been tasked with considering reforms to the criminal law for “egregious regulatory non-compliance” and a standalone “failure to prevent” offence for companies which develop an AI model that engages in harmful conduct.
Prime Minister Anthony Albanese committed to accept all 18 recommendations from the rapid review earlier in November before it was made public.
“Our task is to make sure that we shape AI rather than AI shape us,” Mr Albanese said at the time.
“AI offers Australia and the world generational opportunities if we seize and shape them properly and that is exactly what we’re doing here.”
The release comes as the domestic creative sector has urged the government to uphold copyright laws to prevent Australian works being used to train AI models without consent or compensation.












English (US) ·
Turkish (TR) ·