What is DeepSeek, and why did its cheap AI model shake the stock market?

DeepSeek is a Chinese AI company, based in Hangzhou and owned and funded by the Chinese hedge fund High-Flyer, that builds open-weight large language models. It was founded in July 2023 by Liang Wenfeng. In January 2025 it launched a chatbot alongside its DeepSeek-R1 model, and the app's sudden popularity, together with its performance at relatively minimal cost, is why it made headlines worldwide.

The stock market noticed first. By 27 January 2025, DeepSeek-R1 had surpassed ChatGPT as the most-downloaded free app on the iOS App Store in the United States, and shares in major technology companies plunged that day: Nvidia's stock fell by as much as 17 to 18 percent. In January 2025 Nvidia saw the largest one-day loss in market capitalization for a US company in history, about $600 billion. Publications such as The Hill and The Guardian called the release a Sputnik moment for American AI.

Not everyone agreed on what it meant. Microsoft CEO Satya Nadella and OpenAI CEO Sam Altman called DeepSeek super impressive. Other AI leaders expressed skepticism of its performance or of the sustainability of its success, and Scale AI's Alexandr Wang speculated that China may possess more Nvidia H100 chips than thought.

Governments reacted too. In February 2025 Australia, South Korea and Canada banned DeepSeek from government devices. A US Department of Homeland Security order, made public in the Federal Register on 20 October 2026, excludes DeepSeek products from civilian federal agencies' purchases and orders them removed from their IT systems; it says DeepSeek's models store data on servers in China. Chinese officials criticised the move as using a pretext of national security to harm the commercial interests of Chinese companies.