China Weighs Tighter Controls on DeepSeek and Moonshot AI Models

Beijing is shifting its policy as Chinese open-source AI models gain popularity in Silicon Valley. Models led by DeepSeek and Moonshot AI have become popular low-cost alternatives for Silicon Valley companies because of their exceptionally strong price-to-performance ratio. To prevent core technologies from leaking abroad and models from being maliciously exploited by foreign actors, the Chinese government has launched national security reviews and is considering tighter controls on domestic AI laboratories.

Bloomberg reported in July 2026 that Chinese authorities had begun discussions with DeepSeek and Moonshot AI about imposing stricter security reviews before models are released. If a model is deemed to contain sensitive technology, authorities could restrict access for foreign users or require its release to be delayed. DeepSeek also recently completed a $7.4 billion funding round and is in talks to raise another 10 billion yuan at a valuation of $71 billion.