Silicon Valley Is Completely Divided Over Chinese AI - WIRED

What the report says

Wired reports that Silicon Valley is split over how the United States should respond to increasingly capable Chinese artificial intelligence models, especially “open-weight” systems whose core components are publicly available for adaptation. The debate overlaps with discussions inside the Trump administration about whether and how to restrict Chinese AI tools that can spread quickly through developer platforms, cloud services and local deployments.

According to Wired, larger AI labs have raised security and intellectual property concerns. Anthropic has accused Alibaba of improperly using distillation, a technique in which a smaller model is trained on outputs from a stronger one, and the White House has said it believes Moonshot AI’s Kimi K3 model was developed by distilling Anthropic’s Fable 5. Anthropic CEO Dario Amodei has also warned that open-weight large language models can be downloaded and modified for harmful purposes.

Smaller startups and investors are pushing back against broad restrictions. Wired says more than 200 startups organized as the Little Tech Association, including Y Combinator, wrote to White House science adviser Michael Kratsios and Commerce Secretary Howard Lutnick opposing an outright ban on open-weight models. Their argument is that access to foreign and open models helps US startups compete, lowers costs and prevents dominant AI companies from tightening control over the market.

The issue matters because it pits national security and safety concerns against competition, research access and startup economics. Wired notes that some investors, including Bill Gurley and Chamath Palihapitiya, have publicly criticized efforts that they view as protecting leading AI labs’ business interests rather than the broader technology ecosystem.