US bans Chinese AI startup DeepSeek from federal systems, Beijing pushes back

The U.S. government has ordered civilian federal agencies to stop using products from Chinese AI startup DeepSeek, citing national security risks related to data access and potential espionage. The move reflects growing concern in Washington that not only hardware like chips and network gear, but also AI software itself, can become a strategic vulnerability.

DeepSeek is one of China’s fastest‑rising large language model providers, offering low‑cost, high‑performance chatbots and developer APIs aimed at global customers. U.S. officials worry that sensitive text, code, and other information passed through its systems could ultimately be accessible to the Chinese state, even if indirectly.

Beijing has sharply criticized the ban. China’s foreign ministry accused Washington of “politicizing technology” and distorting fair competition, and warned that countermeasures are on the table. The clash underscores how AI models are becoming entangled in the broader tech and trade conflict between the two countries. For companies and researchers worldwide, it is another signal that geopolitical risk now matters when choosing which AI providers to build on.