Anthropic, OpenAI Push to Restrict Chinese Open-Source AI
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Anthropic and OpenAI have lobbied Washington regulators to restrict Chinese open-source AI models, drawing in Treasury Secretary Scott Bessent and Trump's science adviser Michael Kratsios, who weighed in on the value of American AI models as intellectual property.
- Jensen Huang posted on X for the first time ever saying "the world needs both frontier closed models and frontier open models," with Satya Nadella posting nine minutes later that open-source software was "essential to a healthy A.I. ecosystem."
- Executives from Microsoft, Nvidia, Meta, Palantir, and IBM all signed a letter supporting open-source AI development.
- Z.ai and Moonshot AI, two Chinese AI startups, recently released models that rival those from Anthropic and other American labs, with Anthropic and OpenAI claiming the Chinese firms improperly harvested data from their systems to build them.
- Microsoft and Nvidia rely on open-source models to spur demand for their cloud services and chips, and argue that open development lets more people examine AI security.
- Silicon Valley startup founders argue that constraining open-source software could cement the lead OpenAI and Anthropic hold and unfairly stifle competing companies.
Why it matters: The fight pits two frontier-model labs against the rest of Silicon Valley over whether the federal government should restrict Chinese open-source AI, with lobbying reaching Treasury Secretary Bessent and Trump's science adviser Kratsios. The outcome decides whether Anthropic and OpenAI's market lead gets locked in by policy or whether open-source competition — including cheaper Chinese rivals — continues unchecked.
