Chinese AI Models Still Allowed to Run on Australian Government Systems
Home Affairs officials could not rule out the possibility of government agencies using Chinese artificial intelligence (AI) models during the Joint Select Committee on Artificial Intelligence on Oct. 30.
TL;DR
- Home Affairs officials acknowledged that Australian government agencies might still use Chinese AI models.
- Open-weight AI models can be downloaded and run locally, unlike those accessed through developer websites.
- A government directive from February last year ordered the removal of specific Chinese AI products from government systems.
- Exceptions to the directive are possible if agencies use certified secure environments and implement suitable mitigations, along with having the necessary skills.
- Chinese AI models like Kimi and Qwen have also been scrutinized in the US for potential security risks in their code output.
- The department is developing guidelines for the safe use of open-weight models in critical infrastructure and businesses.
- Australia is pursuing digital resiliency agreements with companies like Microsoft to ensure service continuity during crises.