Home · Technology · Oct 26 archive
US banning Chinese AI DeepSeek from government agency IT systems
Confirmed
In Short: The US Department of Homeland Security has ordered federal civilian agencies to remove all products affiliated with the Chinese artificial intelligence company DeepSeek from their IT systems, citing potential national security risks and concerns over the exposure of sensitive information to Chinese authorities.
The order, issued on July 24 but made public only this week, instructs federal civilian agencies to remove any DeepSeek-related technology, including third-party products or services built on DeepSeek’s AI models.
DeepSeek founder Liang Wenfeng is expected to commit nearly $2.96 billion of his own capital, and the proposed deal is likely to value DeepSeek between $52 billion and $59 billion.
According to the Department of Homeland Security, the removal of DeepSeek products may result in the transfer of sensitive data to the People’s Republic of China if used by U.S. government employees or vendors.
DeepSeek has been accused of having its AI models essentially copied by some Chinese AI firms, including itself, by rivals such as Anthropic and OpenAI.
DeepSeek announced a collaboration with Huawei on September 30, revealing that it is releasing open-source software infrastructure designed specifically for Huawei’s AI processors.
DeepSeek’s collaboration with Huawei focused on improving both computation and communication within the 128-chip system, optimizing specifically for Huawei’s Ascend 950 chips and supporting a “supernode” configuration linking up to 128 of those accelerators.
The US ban on DeepSeek has drawn criticism from China, with the commerce ministry stating that the U.S. is trying to suppress and contain China through discriminatory restrictions under the guise of national security.
China has warned of possible retaliatory measures, highlighting tensions in the tech race between the two countries.
The Department of Homeland Security's order is filed under docket DHS-2026-0628 at regulations.gov, and questions go to a FASCSA implementation team inside the Cybersecurity and Infrastructure Security Agency.
Agencies may submit written exemption requests within 60 days through the Cybersecurity and Infrastructure Security Agency for mission-critical functions.
DeepSeek’s Beam model, a text-only mixture-of-experts model trained on high-compute reinforcement learning, is designed to be effective at reasoning, coding, and agentic tasks at a fraction of the token cost and inference time compute of rivals.
The ban on DeepSeek is part of a broader effort by the US to address concerns over the potential misuse of AI technology by rival firms in China.
Background
The United States Department of Homeland Security has ordered federal civilian agencies to remove Chinese artificial intelligence company DeepSeek’s products and related third-party services from their IT systems, citing concerns that sensitive government information could be transferred to China.
The US has ordered all DeepSeek AI technology from China to be removed from government systems, citing national security risks.
What's confirmed
- The US has ordered all artificial intelligence (AI) technology from the Chinese firm DeepSeek to be removed from most government systems.
- Anthropic and OpenAI, both of which are poised to list on the public stock market in the coming months, have also claimed to have had their AI models essentially copied by some Chinese AI firms, including DeepSeek.
- Federal civilian agency IT systems must remove all products affiliated with the Chinese AI company, the Department of Homeland Security ordered.
What's still developing
- Last year, DeepSeek said it took about $6m (£4.5m) to train the debut version of its model, while the costs associated with OpenAI's GPT-4 were roughly $100m.
- White House tech lead Michael Kratsios wrote earlier this year in an internal memo that "malicious exploitation" of American AI firms had been occurring, mainly by rival firms in China.
- Responding to a Reuters inquiry, spokesperson Mao Ning said China “consistently opposes the over-generalization of national security as well as the politicization of science, technology and trade issues,” adding that Beijing would safeguard the rights and interests of its companies.
- "A global fracture in AI governance would be extraordinarily problematic at a time when the need for global collaboration in this is generally accepted as necessary, if we are to build societal trust in AI," she said.
- Anthropic said it logged more than 12 million exchanges tied to DeepSeek's distillation activity over two weeks in July 2026.
- "could produce actionable intelligence on U.S. persons and/or government agencies. The models themselves may contain malware or other security vulnerabilities."
- Add ANI As A Trusted Source For Reliable Information ANI | Updated: Oct 22, 2026 19:01 IST Washington, DC [US], October 22 (ANI): The United States has directed federal agencies to remove applications developed by Chinese artificial intelligence company DeepSeek from government computer systems, citing potential national security risks and concerns over the exposure of sensitive information to Chinese authorities, The Epoch Times reported.
- According to the report, the federal contracting database SAM.gov lists DeepSeek under an active exclusion effective October 16, without a specified expiration date.
- The US Department of Homeland Security on Monday barred Chinese artificial intelligence firm DeepSeek from federal civilian agency networks, citing espionage and cybersecurity risks amid Washington's intensifying technological rivalry with Beijing.
- DeepSeek, the Chinese AI lab that shook Silicon Valley with its ultra-efficient models, is reversing its long-standing "no funding" policy.
- Reuters noted that rivals have poached researchers from DeepSeek, including Luo Fuli, who left to lead Xiaomi's MiMo model team.