Home Affairs Does Not Rule Out Chinese AI Models

Officials say government agencies could run downloadable Chinese open-weight AI models inside certified secure environments if they can show safe deployment.

Updated on Nov. 3, 2026 in Artificial Intelligence

Home Affairs Does Not Rule Out Chinese AI Models

Live Poll

Should Australian government agencies be allowed to run Chinese AI models?

Home Affairs officials did not rule out government agencies using Chinese AI models when they testified at the Joint Select Committee on Artificial Intelligence on 30 October 2026. Officials said agencies can run downloadable AI models only in certified secure environments.

Why it matters

The question arises because open-weight models can be downloaded and run on a user's own computers, raising security concerns for government and critical infrastructure use across Australia. The department said it was still working out what safe use should look like.

Testimony was given at a Joint Select Committee hearing on 30 October 2026 in response to a question about Kimi and Qwen. The department has not yet defined the standard that safe deployment would require.

The players

Home Affairs

The Australian federal department responsible for national security, border protection and the coordination of cyber and critical infrastructure policy.

Australian Signals Directorate

Australia's foreign signals intelligence and cyber security agency, which agencies must satisfy before deploying such models.

The details

Asked whether Chinese open-weight models could be used by government agencies or critical infrastructure providers, Home Affairs said any use would need to occur inside a protected, certified environment run by a major cloud provider with suitable mitigations in place. Agencies handling classified information must use certified secure cloud environments and show Home Affairs and the Australian Signals Directorate they can deploy such models safely.

30 October 2026: Home Affairs officials testify at the Joint Select Committee on Artificial Intelligence.

Big Picture

The security implications of using Chinese AI models have already been examined in the United States, and Australia's Home Affairs testimony places local deliberations within that wider scrutiny of Chinese-developed open-weight systems.

Australian public servants and contractors will not be able to deploy downloaded Chinese AI models without certified secure environments and proof of safe deployment. Teams using such models outside those conditions face restrictions on their use.

The takeaway

Open-weight AI models can be run on local hardware, so security decisions depend on where and how they are deployed rather than only on their origin. Organisations weighing similar choices should check certification and deployment controls before adoption.