The US government has issued a directive to all federal civilian agencies to immediately cease purchasing and remove all services and products developed by the Chinese AI company DeepSeek from their information systems. This decision is driven by serious national security concerns stemming from DeepSeek's operations in China. Furthermore, the ban not only applies directly to DeepSeek's products but also extends to third-party solutions that utilize its technologies, although an exception can be requested.
The Department of Homeland Security (DHS) specifically expressed concern over DeepSeek's handling of user data. It warned that sensitive information of US government users could potentially be accessible to Chinese authorities in accordance with Chinese laws.
This move is another manifestation of the intense technological rivalry in artificial intelligence between the United States and China. DeepSeek has become a significant market player, competing with American giants like OpenAI and Meta with its low-cost, open-weight models.
China reacted with displeasure to the US measure. Foreign Ministry spokesperson Mao Ning stated that Beijing opposes “exaggerating the concept of national security and politicizing technological and trade issues,” adding that China will “resolutely protect the legitimate rights and interests of Chinese enterprises.”
Planning a new website or app?
Tell me what you have in mind — I will take a look and advise on the best way forward.
The United States had previously taken steps to limit DeepSeek's use in government systems. A recently passed law focused on its use in national security intelligence systems. The latest regulation expands these restrictions to federal civilian agencies, covering both new purchases and existing IT systems.
Meanwhile, according to reports, China is considering implementing export restrictions for 23 AI-related technologies, including chip designs and advanced AI models, with the aim of addressing security and misuse risks.
The Chinese AI lab has been under scrutiny from governments and regulators worldwide since the release of its R1 model, which previously caused significant disruption in the AI industry. Furthermore, in a September report, the United States accused DeepSeek and five other Chinese AI companies, including Alibaba and Moonshot AI, of extensively using model distillation to extract information from American AI systems.
Independently, OpenAI and Anthropic also expressed their concerns about DeepSeek's alleged use of model distillation. OpenAI previously informed US lawmakers that it had observed suspicious attempts by DeepSeek to distill its models, while Anthropic stated that it had identified unauthorized attempts by DeepSeek and other Chinese AI labs to collect outputs from their Claude model. Model distillation is a technique that uses the outputs of one AI model to train another model.