US tech giants are lobbying Donald Trump against a ban on Chinese AI after the launch of a breakthrough new bot last week.
Jensen Huang, the chief executive of Nvidia, joined technology giants including Meta, Microsoft and Palantir in writing an open letter urging the White House not to restrict open-source AI technology, which underpins a powerful new Chinese AI system.
The launch of Kimi K3, an AI bot developed by Chinese start-up Moonshot, has sent shock waves through global markets on fears it could upend the dominance of US AI labs.
The AI bot’s sudden success has prompted calls for restrictions on China, with the White House arguing for sanctions on the country’s AI businesses.
There have also been calls from some US AI companies to restrict open-source, or “open-weight”, AI technology. These systems, which have been heavily relied on by Chinese AI companies, allow the bots to be freely downloaded or altered, and threaten to undercut Silicon Valley giants such as Anthropic and OpenAI.
One senior OpenAI executive last week claimed Beijing’s strategy of giving away its best AI technology amounted to “full AI communism, which is precisely what China proposes”.
However, in a sign of growing alarm over a possible ban in the rest of the tech industry, Mr Huang posted on X for the first time on Friday to oppose the potential restrictions.
Mr Huang said: “AI will transform every industry, power every company, and be built by every country.
“The world needs both frontier closed models and frontier open models.”
The letter from the US technology giants warned that restrictions on more open technology would hinder competition and limit advances in AI.
While more freely available AI would carry risks, such as making AI systems more difficult to trace, the letter said: “The right response to this is not to prohibit open weights.”
The letter, which also received the backing of Elon Musk, warned against “concentrating AI capabilities behind a small number of closed models”. Closed systems include bots such as OpenAI’s ChatGPT and Anthropic’s Claude.
“With the right choices, open weight AI can expand opportunity, strengthen competition, extend American technological leadership, mitigate risk and ensure that the benefits of this extraordinary technology are shared broadly across our economy,” the letter said.
The lobbying comes amid calls for restrictions on Chinese AI models after the launch of the new Kimi system, which is sophisticated enough to rival the most advanced models released by its US rivals.
The White House this week threatened China with sanctions over claims Moonshot had stolen trade secrets from Anthropic to build the system, through a process called “distillation”.
Michael Kratsios, the director of Mr Trump’s office of science and technology policy, accused Moonshot of using distillation against Anthropic’s new, cutting-edge Fable model and claimed China had circumvented restrictions on high-end Nvidia chips.
Mr Kratsios said: “Large-scale, covert industrial distillation aimed at stealing proprietary US technology and undermining American research is unacceptable.”
The letter from Nvidia also warned against potential bans on “distillation”, in which an AI bot harvests information from a rival system, arguing it was a “widely used” development technique.
The launch of Kimi last week prompted a major sell-off across global chipmakers, wiping hundreds of billions of dollars from US stocks.
