White House Shelves Proposed Ban on Chinese AI Models

The White House formally considered imposing sanctions on Chinese open-source AI model developers — primarily Moonshot AI (Kimi K3) and DeepSeek. The measures on the table included trade blacklists and a ban on U.S. cloud companies doing business with those firms. Big Tech pushback proved powerful enough to force the administration to retreat: Nvidia, Microsoft, Google, and Meta all opposed the measures, and Jensen Huang published his first-ever post on X defending open models.

The fault line ran along a surprising divide: OpenAI and Anthropic backed the restrictions, while the largest infrastructure companies stood against them. This signals that U.S. regulation of Chinese AI is moving toward competition rather than outright bans — Chinese open models will continue running on American infrastructure.