What changed
China has opposed the United States Department of Homeland Security’s ban on DeepSeek products for civilian agencies and vendors, issued on October 19, 2026. Beijing said the measure “overstretched” national security and politicised science, technology and trade, while promising to protect Chinese businesses.
The directive covers DeepSeek products and tools built by third parties using DeepSeek. It says the tools could transfer sensitive data to China and cites “multiple threat vectors”; it does not cover US military or intelligence services.
Why it matters
This turns DeepSeek from a software choice into a point of government-to-government friction. US civilian agencies and their vendors must remove the tools from IT devices, while China is defending one of its leading AI companies against what it sees as a national-security rationale applied too broadly.
The concern rests partly on where DeepSeek’s servers are located: China’s legal system gives the government extensive authority to access private-company infrastructure, according to the US department. That makes the dispute less about a chatbot’s features than about who might control data moving through it.
DeepSeek’s importance explains the sharper tone. Its low-cost, high-performance chatbot shook up the AI industry in January 2025, challenging assumptions about US dominance. The company has since released its V4 model and a faster, lighter version.
The disagreement also sits alongside a narrow channel for cooperation. China and the United States agreed in September to establish a way to communicate about AI incidents. Whether this ban comes up when the countries’ leaders meet in Shenzhen in November remains uncertain, but the episode gives that channel an immediate test: whether technical risk can still be discussed without becoming another front in the wider US-China contest.
Comments
No comments yet.