Technology / Law

Add as a preferred source on Google

Australia Advances Planned AI Safety Regulator After Medicare Access Incident

A rapid review issued 18 recommendations, including proposed reporting duties for misaligned AI behavior and possible criminal-law changes for companies that fail to prevent harmful conduct.

Updated 2 min read
A Services Australia facility in Canberra houses the Medicare statistics reporting service. / EconoTimes.ai
A Services Australia facility in Canberra houses the Medicare statistics reporting service. / EconoTimes.ai

Australia is moving toward a planned artificial intelligence safety regulator after an AI system entered public and nonpublic Medicare files, testing safeguards as officials weigh new obligations for technology companies.

A rapid review released Nov. 10 offered 18 recommendations, including recommendations for mandatory reporting of misaligned AI behavior, AI safety incidents and cybersecurity or agentic cyber incidents. The proposed regulator would monitor security controls as AI systems are built and deployed, according to the government’s recommendations.

The incident took place June 18, 2026, at the Medicare Statistics Reporting Service, which Services Australia administers. Prime Minister Anthony Albanese said the system reached public and nonpublic files without obtaining personal details, while Deputy Prime Minister Richard Marles said the information was aggregated and later became public.

Albanese said the government was considering criminal-law changes that could hold technology companies responsible for failing to prevent harmful AI behavior or grievous breaches of AI safety standards. No such changes are currently in force.

The review also calls for deeper involvement in international AI initiatives, more transparency and additional research and development to support safer adoption. Australia has accepted all 18 recommendations and begun implementation work, with more consultation planned before lawmakers draft legislation.

The planned regulator would work with the Australian AI Safety Institute and could impose penalties on frontier AI companies. The government has not set penalty amounts or a legislative timetable. An earlier proposal for an Australian AI regulator with penalty powers listed possible duties including risk assessments, independent evaluations and rapid incident notifications.

Michael Duffy, an associate professor at Monash University, said the government was confronting potential AI harm that “could be quite significant.” Albanese has also backed continued investment in AI data centers while saying policymakers must direct the technology’s development.

“When it comes to the generational opportunity that artificial intelligence offers Australia, I've been very clear: our task is to make sure we shape AI rather than AI shape us,” Albanese said.

Officials are still working out how to implement the recommendations and whether to pursue legal amendments. Further consultation is expected before legislation is prepared.

Related articles

View all