US Treasury Secretary Scott Bessent says the Trump administration will look into whether Chinese AI models were distilled from US models and may “sanction them”
The administration’s review follows a White House claim that China-based foreign entities were conducting “industrial-scale” distillation of American AI technology. The issue has moved from an asserted IP and security concern toward a possible economic-policy response.
It also lands as lawmakers examine US companies’ use of Chinese models and officials reportedly revisit de facto restrictions on foreign open-source models. That makes the Treasury inquiry consequential for both model suppliers and US adopters.
First-order effects
Chinese AI model providers could face a US Treasury review over alleged distillation, with sanctions presented as a possible—not yet enacted—outcome.
US companies using or evaluating those models face added compliance and supplier-risk uncertainty while the administration determines whether to act.
Second-order effects
US AI vendors can use heightened scrutiny of Chinese alternatives to reinforce security and provenance arguments with enterprise and public-sector buyers.
A sanctions path would push procurement, cloud, and model-distribution intermediaries to screen affected providers and reconsider access to their models.
Third-order effects
If allegations of model distillation increasingly trigger financial restrictions, model provenance could become an enforceable trade and national-security issue rather than solely a technical or contractual dispute.
The policy direction points toward more state-mediated segmentation of AI ecosystems, though the scope will depend on what the investigation substantiates and which measures are ultimately adopted.
The trend: AI model access, provenance, and deployment are becoming instruments of US-China economic and security policy.
@MikeIsaac the literal opposite is happening tons of investors have strongly advocated for pro open source position both in america and china https://x.com/... https://x.com/... https://x.com/... https://x.com/...
what will happen for the foreseeable future is basically most of the big labs and VC class lobbying the government against open source chinese models that is b/c open source erodes the first mover advantages closed source companies had and thats where all the investment is
Treasury Sec. Bessent on Fox: “This administration supports open source models, but what we do not support is IP theft. If we see, especially, that overseas models are stealing from our great companies, we have the ability to sanction them because of this theft.”
If there has been “theft” that suggests a crime has been committed. But I am unaware of any lawsuits being filed. No company should be allowed to declare infringement without adjudication. I'm not convinced a court would call using a product as it's designed theft. There is a
“It is going to fail miserably.” I concur. There is zero action DC can take that will stop or slow China's AI progress. In fact, DC action will very likely help them globally. Tyler is spot on.
Bessent lays out the playbook: 1. Declare distillation to be theft. 2. Find that every foreign model has distilled a US model. 3. Sanction those companies and make it harder to use their models. This is a playbook for giving people outside the US, including malicious actors,
U.S. AI policy for the foreseeable future is going to be an open sewer of bizarre, incoherent, xenophobic protectionism, and our trash press is going to normalize it as serious adult policymaking the entire way
New: Undersecretary of War Emil Michael tells me Chinese openweight AI models don't currently need restrictions — “When does it reach the threshold of a national security threat? We're going to have to act, but right now we're not there. We want competition. Competition first.”
translation: it's ok if our companies steal IP wholesale from artists and writes, but bad if people from other countries steal IP wholesale for us. got it!
This from @SecScottBessent is the right approach. We should be going after Chinese open-weight models because they present unique security risks and steal American IP—not necessarily because they are open-source or threaten the position of our leading labs.