美国禁政府系统使用 DeepSeek，中国抗议并指其歧视，但国内仍封锁 ChatGPT、Claude 和 Gemini

The US is banning Deepseek from government systems, and Beijing is calling it discrimination. Yet China locks out foreign AI far more strictly. The protest is still more than political theater.

The US Department of Homeland Security recently ordered civilian federal agencies to remove all products from Chinese AI company Deepseek from their IT systems. The ban also covers software from other vendors built on Deepseek's open-source models, and it prohibits future purchases.

Washington says it's worried Beijing could access sensitive data through Deepseek. China has now pushed back hard. The Foreign Ministry accuses the US of stretching the concept of national security too far, while the Commerce Ministry is demanding the ban be lifted and has vowed to take resolute countermeasures.

On the substance, Beijing is on shaky ground, since China walls itself off far more consistently. ChatGPT, Claude, and Gemini aren't approved there. Since 2023, the Cyberspace Administration has been required to review AI services, and those services must comply with political guidelines. That rule applies to the entire market, not just government agencies. State bodies and state-owned companies are supposed to gradually replace foreign hardware and software with domestic products. Reports also point to restrictions on iPhones for government officials.

According to the NYT, Beijing also proposed its own export and licensing controls for AI technologies last week, similar to the American approach.

China's outrage rings hollow but still sends a political signal

The US concerns also have a concrete basis. China's 2017 National Intelligence Law requires companies and citizens to support the work of the country's intelligence services. Chinese models also carry a political slant, steering clear of certain topics and repeating state propaganda less critically than their Western counterparts. Then there's the economic side. AI companies, the FBI, and other US agencies have repeatedly accused Chinese firms of copying top US models on an industrial scale. Beijing denies this.

Still, the protest serves several purposes. It shows domestic companies that the state has their back. With other countries, it supports the narrative that the US wants to monopolize the AI industry. Most of all, it lays the groundwork for later retaliation, which Beijing can then frame as a mere response.

In the end, both sides invoke national security to protect their markets and technology. The US also ties security and economic policy together through its strict chip export controls. China's outrage isn't very credible, but the political signal behind it deserves to be taken seriously.

AI News Without the Hype – Curated by Humans

Subscribe to THE DECODER for ad-free reading, a weekly AI newsletter, our exclusive "AI Radar" frontier report six times a year, full archive access, and access to our comment section.