Tech Giants Warn Washington Against Rushing to Restrict Open AI Models

A coalition of major U.S. technology companies, including Nvidia, Microsoft, Meta, and Palantir, sent a letter to policymakers on Friday urging them to avoid what they called "premature restrictions" on open-weight artificial intelligence models. The group, which comprises more than 20 firms, argued that such restrictions could "stifle competition or drive innovation overseas." Open-weight models, which allow users to download, modify, and run AI software on their own systems, have become a flashpoint in the industry as Chinese rivals like Moonshot AI release models that outperform leading American offerings on key benchmarks.

The letter comes amid growing concern in Washington about the rise of Chinese open-weight models. Moonshot AI's Kimi K3 model recently beat cutting-edge U.S. systems on some tests, and Treasury Secretary Scott Bessent told CNBC the Trump administration is examining whether Chinese firms have stolen American intellectual property. But the U.S. tech companies warned against any rash moves, writing that open-weight models strengthen competition and ensure that AI's benefits are "broadly shared rather than concentrated in a few hands." They also cautioned that relying solely on closed models, like those from OpenAI and Anthropic, is not inherently safer - those models can be breached or misused in ways outsiders cannot detect.

Notably, OpenAI and Anthropic, both valued at nearly $1 trillion and preparing for potential IPOs, did not sign the letter. However, OpenAI President Greg Brockman said Thursday that the company supports broad access to AI and has not been involved in any talks about banning Chinese models. White House advisor Michael Kratsios claimed that Moonshot AI developed its model by distilling Anthropic's technology - a training method that can be used legitimately or, as he warned, for covert theft. The tech coalition urged policymakers to address unlawful distillation through targeted legal and commercial frameworks rather than sweeping restrictions. "Our AI leadership will be judged not by one frontier AI model, but by whether the United States builds a strong, open ecosystem," the letter concluded.

Source: CNBC