Chinese AI models at the centre of growing divide in Silicon Valley
US Treasury Secretary Scott Bessent has publicly questioned whether Chinese developers are “using stolen materials”, while Beijing has pushed back against American threats of restrictions.
Some in the US tech industry argue that restricting access to Chinese AI could do more harm than good.
New: You can now listen to articles.
This audio is generated by an AI tool.
Read a summary of this article on FAST.
Get bite-sized news via a new
cards interface. Give it a try.
Click here to return to FAST
Tap here to return to FAST
FAST
SAN FRANCISCO: The technology rivalry between the United States and China has opened a new front over artificial intelligence, as Washington considers restrictions targeting Chinese AI companies and their models.
AI models such as DeepSeek, Alibaba’s Qwen, Z.ai’s GLM and Moonshot AI’s Kimi have been thrust into the centre of a growing divide in Silicon Valley.
As Chinese AI models gain popularity in the tech industry, US officials have questioned whether some were developed using technology copied from American firms.
Treasury Secretary Scott Bessent in July threatened sanctions against China, amid reports the US is also weighing a ban or curbs on foreign-made open-weight models that are often built through distillation.
Distillation is a common technique in which one AI model learns from another.
The method is widely used across the industry, but US officials say Chinese firms have deployed it at scale to illicitly copy proprietary American systems.
Beijing has pushed back against American threats of restrictions.
China’s Ministry of Commerce said the US accusations “lack factual basis and legal grounds”.
“China will take all necessary measures to firmly safeguard its legitimate and lawful rights and interests,” it added.
As the Trump administration weighs sanctions and other trade restrictions against Chinese AI firms, the prospect of limiting access to their models is dividing the US technology industry.
Guess Word
Crack the word, one row at a time
Buzzword
Create words using the given letters
Mini Sudoku
Tiny puzzle, mighty brain teaser
Mini Crossword
Small grid, big challenge
Word Search
Spot as many words as you can
IMPACT ON STARTUPS
At software development platform Replit, engineers move seamlessly between US models such as Anthropic’s Claude and Chinese open-weight models including GLM.
Replit CEO and co-founder Amjad Masad said the technology should not be viewed simply in terms of national origin.
“It’s important that we think of it as an ecosystem and not, this is a Chinese model, this is an American model,” he said.
The view is shared by some groups representing smaller technology companies.
Last month, the Little Tech Association urged the Trump administration not to block access to Chinese open-weight AI models.
“(It) would channel all of the demand for AI into a handful of closed-weight companies. That will absolutely escalate prices,” said Harry Godfrey, the industry group’s executive director.
“It will also reduce the ability to actually innovate and build new and novel products.”
WHEN DOES AI LEARNING BECOME IP THEFT?
The latest US allegations have focused on Moonshot AI, the Chinese company behind the Kimi models.
US officials have accused the firm of improperly distilling one of Anthropic’s most advanced AI models to create its Kimi K3.
“If you’re one of these companies that made the models (that) are being stolen, you (will) have a huge problem with this,” said Pedro Domingos, professor emeritus of computer science and engineering at the University of Washington.
For the US government, protecting American companies could be a response to what amounts to “intellectual property theft on some level”, said Domingos.
However, exactly when AI model distillation crosses the line into intellectual property theft remains difficult to determine, leaving policymakers and technology companies grappling with where the boundaries lie.
Some in the US tech industry argue that restricting access to Chinese AI could do more harm than good.
“It would definitely put American companies at a disadvantage,” said Replit’s Masad.
“As an entrepreneur, you want to think really hard about whether it makes sense for you to be based in the US.”
The dispute presents Washington with a difficult balancing act: protecting US companies and their technology while ensuring that startups still can continue to access the technology they need to remain competitive.
As the US-China AI rivalry heats up, Silicon Valley’s concern is whether such measures could ultimately end up driving innovation – and innovators – elsewhere.
Source: CNA/oc(mp)
Sign up for our newsletters
Get the CNA app
Stay updated with notifications for breaking news and our best stories
Get WhatsApp alerts
Join our channel for the top reads for the day on your preferred chat app
Get bite-sized news via a new
cards interface. Give it a try.
Click here to return to FAST
Tap here to return to FAST
FAST
About The Author
