Australia banned one Chinese AI app from government devices last year. The question now is whether other Chinese technology could still slip in.
Officials from Australia’s Department of Home Affairs could not rule out government agencies using Chinese AI models during a parliamentary hearing on Friday, Oct. 30, 2026. The exchange came during a meeting of the Joint Select Committee on Artificial Intelligence in Canberra.
What did officials say about Chinese AI models?
Independent lawmaker Kate Chaney, who represents the Perth area seat of Curtin, asked whether Chinese models that can be downloaded and run on local computers could be used by government agencies or critical infrastructure providers right now. Chaney is one of the independent members of Parliament known as teals, who campaign on climate, integrity and good governance.
A senior Home Affairs official responsible for protective security said rules have applied to federal agencies since the government’s DeepSeek directive in February 2025.
He described those rules in broad terms. Agencies must have the skills to deploy such models safely, must use certified secure cloud systems when classified information is involved and must show Home Affairs and the Australian Signals Directorate, the government’s cyber intelligence agency, that they can meet those standards.
What was the DeepSeek directive?
The directive ordered agencies to block and remove products, apps and web services from the Chinese company DeepSeek from all government systems and devices, following a threat and risk assessment. The ban applied across federal departments and agencies, covering both work devices and systems.
DeepSeek drew global attention in early 2025 when it released a powerful AI model that rivaled American systems at a fraction of the cost. Its release shook global technology markets and raised questions about the U.S. lead in AI.
Several other governments, including those in Italy, South Korea and Taiwan, also restricted DeepSeek over security and privacy concerns.
What are downloadable Chinese AI models?
Some AI developers release their models so anyone can download and run them on their own computers, rather than accessing them through the company’s website or app. These models are released with their underlying files, known as weights, made public.
Several leading Chinese AI models, including versions from DeepSeek and Alibaba, are distributed this way, making them free and easy to adopt. Businesses and researchers around the world have embraced them because they can be customized and run without paying fees.
Running a model on local systems can limit how much data is sent back to the developer, but security experts warn that Chinese AI models may still carry hidden risks, including censorship built into their answers or vulnerabilities that are hard to detect.
Why are lawmakers concerned?
China’s laws can require companies to cooperate with state intelligence work, raising concerns about how Chinese technology could be used to access sensitive information. Australia has previously restricted Chinese technology, including barring Huawei from its 5G network in 2018.
Critical infrastructure, including energy, water and telecommunications, is a particular focus because a breach involving Chinese AI models could affect millions of Australians.
The hearing follows earlier warnings in Parliament about outdated government technology systems and a data breach involving Medicare information.
What comes next for Chinese AI models in government?
The committee is examining how Australia should manage the risks and benefits of artificial intelligence, and its findings could lead to tighter rules. Lawmakers from across the political spectrum have pushed for clearer standards on AI use in government.
For now, officials say agencies wanting to use Chinese AI models must prove they can do so safely, though the hearing showed that no blanket ban exists.
