Australia is moving to establish a dedicated artificial intelligence safety regulator and implement mandatory industry safeguards. Prime Minister Anthony Albanese announced the policy shift following a rapid review into a June security breach involving a government portal by OpenAI.
While global anxiety grows over advanced technology following security incidents involving major firms like OpenAI and Anthropic, Australia is pushing forward with stricter oversight. Although OpenAI will not face criminal penalties over the specific Medicare portal hack, the government is introducing sweeping regulatory changes to reshape how artificial intelligence developers operate within the country.
Establishing an AI Safety Regulator and Mandatory Safeguards
The newly announced safety regulator will be tasked with keeping the booming AI industry in check and ensuring advanced models are independently evaluated. Prime Minister Albanese emphasized that the goal is to shape technology in the national interest rather than letting it dictate terms.
The government outlined six key obligations that artificial intelligence companies will be expected to follow:
- Maintaining security controls across all stages of development and deployment.
- Identifying and managing serious risks associated with their models.
- Providing rapid, mandatory notification of security incidents to the Australian government.
- Allowing independent evaluation of frontier AI models.
- Maintaining and publishing safety and security frameworks to ensure transparency, good governance, and risk management.
- Responding promptly to updated standards and rules as technology evolves.
Legal Reform and Industry Backlash
Alongside the creation of the safety regulator, the Australian government is considering criminal law reforms. Recommendations from the rapid review—conducted by the attorney general’s department, the Australian federal police, and home affairs—suggest introducing penalties for egregious regulatory non-compliance. This includes a potential "failure to prevent" offense that would criminalize developing an AI model that engages in harmful conduct, closing legal loopholes where companies might argue a computer's actions were unexpected.
The tightening of rules comes amid a broader local backlash against AI developers in Australia. While the country has positioned itself as an attractive hub for power-hungry data centers—including an OpenAI project on the outskirts of Sydney—local communities have raised concerns over ethics, environmental footprints, and whether copyright laws will be weakened to allow chatbots to train on books, songs, and artwork.
