The short version
- DHS banned all use of DeepSeek AI and technologies built with its models in civilian agencies like Treasury, State, and Justice, effective from late July.
- The move follows White House criticism of Chinese AI firms for 'malicious exploitation' and claims by US companies that their models were copied by rivals including DeepSeek.
- DeepSeek remains popular globally for its low-cost, open-source models, but the ban excludes Department of Defense agencies and reflects broader US efforts to curb Chinese influence in critical tech.
The United States has ordered the removal of artificial intelligence technology from the Chinese firm DeepSeek from most government systems, marking a significant escalation in the ongoing tech rivalry between Washington and Beijing. The Department of Homeland Security (DHS) issued a ban prohibiting all use of DeepSeek within civilian government agencies, as well as any technology developed with the aid of its AI models. Although the order was made public on Monday, it has been in effect since the end of July.
The ban applies to major civilian departments including the Department of the Treasury, the Department of State, and the Department of Justice. However, agencies within the Department of Defense are explicitly excluded from this specific directive. The DHS notice did not detail the exact security concerns surrounding DeepSeek, but the move aligns with broader White House criticism of Chinese AI firms throughout the year.
White House technology lead Michael Kratsios wrote in an internal memo earlier this year that 'malicious exploitation' of American AI firms had been occurring, primarily by rival companies in China. This sentiment is echoed by US tech giants Anthropic and OpenAI, which have claimed that their AI models were essentially copied by some Chinese firms, including DeepSeek. Both companies are poised to list on the public stock market in the coming months.
President Donald Trump has consistently identified China as the key rival in the development of AI, viewing the technology as crucial to the future of the United States. The ban on DeepSeek reflects this strategic priority, aiming to prevent potential security vulnerabilities associated with foreign-controlled AI infrastructure within sensitive government operations. The DHS, DeepSeek, and the Chinese embassy in Washington DC have been contacted for comment regarding the order.
DeepSeek has gained significant popularity since its initial public release in 2025, largely due to its open-source nature and lower costs compared to closed models from US competitors. Its free chatbot app is among the most downloaded on the Apple App Store, and its AI models are integrated into offerings by US tech companies such as Snowflake, Vercel, and OpenRouter. Startups have flocked to DeepSeek because using its models costs a fraction of what it would to use tools like Google's Gemini, Anthropic's Claude, or OpenAI's Codex.
The cost disparity is stark: DeepSeek reported spending approximately $6 million to train the debut version of its model last year, whereas OpenAI spent roughly $100 million on GPT-4. These revelations have sent tech stocks tumbling and raised concerns about the sustainability of US AI dominance if Chinese firms can produce competitive models at a fraction of the cost. The most recent version of DeepSeek was released last week, with users praising its performance in common tasks such as software coding.
Despite the ban on federal civilian agencies, DeepSeek's influence in the broader tech ecosystem remains substantial. Its servers have occasionally gone down due to high demand, indicating widespread adoption. The US government's move to exclude it from civilian systems is part of a larger effort to secure critical infrastructure against potential espionage or manipulation by foreign adversaries.
The exclusion of Department of Defense agencies from this specific ban suggests that the military may already have separate protocols or restrictions in place regarding Chinese AI technology. Alternatively, it could indicate a different risk assessment for defense-related applications compared to civilian government functions. The distinction highlights the nuanced approach the US is taking in managing its technological competition with China.
As the race for AI dominance continues, the ban on DeepSeek underscores the growing intersection of technology and national security. While the immediate impact is limited to federal civilian agencies, the decision sends a strong signal to the private sector about the risks associated with relying on Chinese AI infrastructure. The long-term implications for global AI development and US-China relations remain to be seen.
What remains unresolved is whether this ban will lead to broader restrictions on other Chinese tech firms or if it will spur increased investment in domestic AI alternatives. The White House has not specified if similar measures will be applied to other entities, but the precedent set by the DHS order could influence future policy decisions. Meanwhile, DeepSeek continues to operate globally, maintaining its position as a disruptive force in the AI landscape.
Sources behind this briefing
Go to the original reporting
- BBC Technology↗US bans China's DeepSeek from government in race for AI dominance
- Reuters↗US banning Chinese AI DeepSeek from government agency IT systems
- BBC↗US bans DeepSeek AI from government use as tech tension with China continues
