Australia bans DeepSeek on govt devices citing security concerns

Australia has officially prohibited the use of the DeepSeek chatbot on all government devices after security agencies raised concerns over potential privacy and malware risks linked to the AI program developed by a China-based startup.

DeepSeek, which launched last month, quickly gained attention and disrupted financial markets. However, worries over its security and data handling have prompted several nations, including South Korea, Italy, and France, to voice similar concerns.

Australia’s decision stands out as one of the strongest responses to the Chinese-developed chatbot so far. Government cybersecurity envoy Andrew Charlton stressed the gravity of the move, stating, "This is not a symbolic gesture but a measure based on advice from security agencies. We cannot risk exposing government systems to these applications."

Charlton explained the risks, noting that information uploaded to DeepSeek "might not remain private" and that the application could "expose users to malware."

An 'unacceptable' security risk
The Department of Home Affairs issued a directive to government staff, with Secretary Stephanie Foster outlining the risk posed by DeepSeek: "After reviewing the threat and risk analysis, I have concluded that DeepSeek's products, applications, and web services present an unacceptable security risk to the Australian Government."

Effective immediately, all non-corporate Commonwealth entities are required to "identify and remove all instances of DeepSeek products, applications, and web services from government systems and mobile devices," Foster added. The directive also mandates preventing "access, use, or installation" of DeepSeek across government platforms.

The ban has received bipartisan support from Australian politicians. This decision mirrors past actions, such as Australia's 2018 ban on Huawei from its 5G network over national security concerns and the 2023 ban on TikTok from government devices based on advice from intelligence agencies.

Cybersecurity researcher Dana Mckay emphasized the risks, stating, "Chinese companies are required to store data in China, making it subject to government inspection." Mckay, from the Royal Melbourne Institute of Technology, also pointed out that DeepSeek's privacy policy explicitly mentions it collects keystroke data, which could be used to identify individuals.

"With that data, you could leverage sensitive information if you knew it originated from a government machine," Mckay warned.

Alarm bells raised
DeepSeek caused alarm last month when it claimed its new R1 chatbot could match the capabilities of leading AI systems in the US at a fraction of the cost. This has led to a stir in Silicon Valley, with some experts calling it a wake-up call for US developers.

There are also accusations that DeepSeek reverse-engineered US-developed AI, like that used in ChatGPT. Countries including South Korea, Ireland, France, Australia, and Italy have all voiced concerns about DeepSeek's data handling, particularly around the use of personal information and training data for its AI system.

Tech and trade disputes between China and Australia have simmered for years. Beijing was infuriated by Australia’s decision to ban Huawei, as well as its crackdowns on Chinese foreign influence and calls for an investigation into COVID-19’s origins. A trade war between the two nations ensued but cooled down last year when China lifted its last trade barrier, a ban on Australian live rock lobsters.