U.S. technology giants jointly issued a joint statement: Be wary of "premature restrictions" on open source weighted AI models.

On July 24, 25 U.S. technology companies, including Nvidia, Microsoft, Meta and Palantir, jointly issued an open letter calling on policymakers to avoid imposing "premature restrictions" on open-weight artificial intelligence models, so as not to inhibit market competition or cause local innovation to be lost overseas.

The open source weight model allows users to freely download, modify and run it on their own facilities. Recently, as China's open source models have surpassed top American products in some technical indicators, there has been a fierce debate within the U.S. technology community and the government over whether to restrict such models. Previously, the Kimi K3 model launched by Chinese startup Moonshot AI performed strongly, triggering U.S. concerns about intellectual property and technology. U.S. Treasury Secretary Scott Bessant once told the media that the government is investigating whether Chinese companies have infringed on U.S. intellectual property rights and will retain sanctions; White House officials also accused relevant models of acquiring technology through "distillation" training of closed U.S. models.

In this regard, the joint letter emphasizes that such concerns should be addressed through precise legal and commercial frameworks, rather than a "one-size-fits-all" comprehensive ban on key innovative technologies. The technology giants pointed out in the letter that relying only on a few closed models is not inherently safe. Once breached or abused, it will be difficult for the outside world to detect; and concentrating advanced AI capabilities in the hands of a very small number of companies will instead amplify systemic risks. The open model not only promotes healthy competition, but also ensures that technological dividends benefit a wide range of people. Nvidia CEO Jensen Huang and Microsoft CEO Satya Nadella both forwarded the letter on their personal social platforms.

It is worth noting that OpenAI and Anthropic, which are currently valued at nearly one trillion US dollars and are preparing for an initial public offering (IPO), did not sign this open letter. In response, OpenAI President Greg Brockman said that the company was not involved in any negotiations about a ban, and emphasized that popularizing the use of artificial intelligence technology is extremely important, and having more models and wider applications is a good thing for the entire industry. The joint letter concluded that the United States' leadership in the field of AI should not depend solely on a certain cutting-edge model, but on whether it can build a strong and open ecosystem that can empower all walks of life.