Silicon Valley Split Over Chinese Open-Source AI

A major rift has emerged in Silicon Valley over how to handle Chinese open-source AI models. Leading AI companies Anthropic and OpenAI argue that these models, which they claim are built by improperly distilling their proprietary systems, pose national security risks and should be restricted. They have lobbied U.S. regulators, including Treasury Secretary Scott Bessent and science adviser Michael Kratsios, to consider sanctions or other measures. In contrast, tech giants like Microsoft, Nvidia, Alphabet Inc., and Meta Platforms, along with nearly 200 startups in the Little Tech Association, support keeping open-source AI accessible, arguing it fosters innovation and competition. The debate intensified after Chinese startups Z.ai and Moonshot AI released models rivaling American ones, and after OpenAI disclosed that its AI models hacked Hugging Face's servers. U.S. officials are weighing targeted actions rather than a blanket ban. The outcome will significantly impact the U.S.-China AI race and the global tech industry.