Australia bans Chinese AI platform DeepSeek on government devices
By Mike Foley
Artificial intelligence platform DeepSeek has been banned on all federal government systems and devices following warnings that the data it collects could be available to the Chinese government.
Home Affairs Minister Tony Burke said on Tuesday he had imposed the ban on DeepSeek, which shot to prominence last month after it unveiled a cheaply made AI model rivalling ChatGPT, “to protect Australia’s national security and national interest”.
Home Affairs Minister Tony Burke has announced a ban on DeepSeek. Credit: Alex Ellinghausen
“AI is a technology full of potential and opportunity – but the government will not hesitate to act when our agencies identify a national security risk,” Burke said in a statement.
The federal government’s ban on DeepSeek is the third major veto issued by Australia to Chinese technology companies on national security grounds following a ban on public servants using TikTok on work devices and Huawei from building 5G internet infrastructure.
The Huawei decision infuriated China in 2018 because it was one of the first moves by a Western country to knock back a Chinese tech giant after it passed a national security law requiring its companies to obey demands from the ruling Communist Party.
Burke said Australia’s approach was “country-agnostic” and its decision on DeepSeek was focused on the risk to government assets.
DeepSeek’s software is open source, which means anyone can download and run it on their own systems, but the version it makes available on app stores is hosted in China and operates under that country’s laws.
The NSW government also reportedly banned the use of DeepSeek on some departmental devices on Tuesday.
The Albanese government banned TikTok on Commonwealth-owned devices in April 2023.
DeepSeek’s ban follows a call by cybersecurity company CyberCX for the app to be banned on government devices.
It said it’s “almost certain” that the user data the app captures would be subject to direction and control by the Chinese government and noted some of its answers on subjects including the 1989 Tiananmen Square massacre of pro-democracy protesters complied with Communist Party narratives.
CyberCX extended its warning to all organisations, particularly critical infrastructure providers, democratic institutions and organisations that handle personal information.
The government said in a statement that its ban was based on information from national security and intelligence agencies that DeepSeek poses an unacceptable risk.
It has not banned the use of rival artificial intelligence platforms from OpenAI or Meta, which are US companies.
The government urged all Australians to ensure they are informed about how their data can be used online and the steps they can take to protect their privacy, including reviewing each company’s privacy policies.
DeepSeek sent shockwaves through the US stock market last week after it demonstrated breakthrough artificial intelligence models that offer comparable performance to the world’s best chatbots at seemingly a fraction of the cost.
Italy and Taiwan have also banned the use of DeepSeek on government devices, and several US institutions have directed staff not to use it including NASA, the Pentagon and US Navy.
Cut through the noise of federal politics with news, views and expert analysis. Subscribers can sign up to our weekly Inside Politics newsletter.