Silicon Valley Split as OpenAI, Anthropic Push to Block Chinese AI Models
Leading U.S. AI labs lobby for restrictions on open-weight Chinese models while startups warn the move would cement incumbent power and stifle innovation.
Silicon Valley is fracturing over whether the United States should restrict access to Chinese open-weight AI models—a question that could define the future of artificial intelligence.
OpenAI and Anthropic have lobbied Washington regulators to impose limits on Chinese open-weight systems, according to The New York Times, which cited five people close to the discussions. Opposing them: a coalition of nearly 200 startups, organized through the Little Tech Association, urging the Trump administration against blanket bans they say would hand market advantage to a few closed-source incumbents.
Chinese Models Close the Gap
The debate has intensified following the release of high-performing Chinese models that rival top U.S. systems. On July 16, 2026, Moonshot AI unveiled Kimi K3, a 2.8-trillion-parameter open-weight model that ranks first on Arena Frontend Code and third or fourth on independent intelligence indices. Weeks earlier, on June 13, Z.ai released GLM-5.2, a 753-billion-parameter model offering a 1-million-token context window.
Both models are positioned as lower-cost alternatives to U.S. frontier systems, enabling rapid integration into global developer workflows. China has aggressively pursued open-source and open-weight strategies even as the U.S. maintains hardware restrictions on advanced chip exports.
Anthropic Bets on Regulation
Anthropic has put significant financial backing behind its regulatory stance. On February 12, 2026, the company announced a $20 million donation to Public First Action, a bipartisan 501(c)(4) organization that advocates for AI regulation and plans to support 30 to 50 candidates in state and federal races. Anthropic later committed an additional $20 million, bringing total support to $40 million.
Open vs. Closed Future
The outcome of this regulatory battle will determine whether AI development remains distributed and collaborative or consolidates under a few massive corporations with government backing. Restricting Chinese open-weight models could protect revenue streams for OpenAI and Anthropic but risks alienating developers who rely on low-cost, high-performance alternatives.
There is also the specter of retaliation: Beijing could impose reciprocal restrictions on U.S. models, further fragmenting the global AI ecosystem. For now, the industry watches Washington, where the definition of "fair competition" in AI is being written behind closed doors.