Following a security breach involving OpenAI and Australia's Medicare system in June, the Australian government has announced plans to introduce mandatory incident reporting for artificial intelligence developers. Prime Minister Anthony Albanese released the recommendations of a rapid review into the breach on Tuesday, signaling stricter oversight to prevent future security failures.
While OpenAI will not face criminal penalties over the June incident, the government is signaling upcoming changes to criminal law. The proposed updates aim to ensure technology companies cannot evade liability by claiming ignorance of how their automated systems operate.
Under the newly outlined measures, AI developers operating in Australia will be required to:
- Maintain robust security controls throughout development and deployment.
- Identify and actively manage serious risks tied to their models.
- Provide rapid, mandatory notifications to the Australian government following security incidents.
- Permit independent evaluations of their AI models.
- Publish and maintain safety and security frameworks focused on transparency and governance.
- Adapt to evolving standards and regulatory rules as technology advances.
The federal government intends to use these updated standards and legal frameworks to combat potential risks as artificial intelligence adoption grows.
