![US Treasury Secretary Scott Bessent attends a meeting between President Donald Trump and Lebanese President Joseph Aoun in the Oval Office at the White House in Washington on Tuesday. [EPA]](https://wimg.heraldcorp.com/news/cms/2026/07/22/news-p.v1.20260722.6dd7d77a3e08472a88a9e859bf3d0c24_P1.jpg)
US Treasury Secretary Scott Bessent said Tuesday that the government would scrutinize Chinese open-source AI models and take action if evidence emerges that they have stolen intellectual property from American companies.
Speaking to Fox Business, Bessent said the administration had heard extensive discussion about open-source models threatening US AI models. "The government supports open-source models," he said. "But what we do not support is intellectual property theft."
He went on to say that if foreign models were found to be stealing the fruits of American companies' work, "we have the authority to sanction them."
Bessent said Chinese models had sent "hundreds of millions of calls to US models in an attempt to reconstruct the model's code and processes," calling it "unacceptable" and pledging a thorough investigation "within days or weeks."
His remarks echo claims by OpenAI and Anthropic that Chinese rivals have systematically and without authorization copied the outputs of leading US models to develop a new generation of chatbots at far lower cost.
In April, the Trump administration pledged further measures — including steps to hold foreign actors accountable — over the practice known as adversarial distillation.
Bessent also raised the question of whether US companies using Chinese AI models should be required to disclose that fact to their customers, stressing that consumers "should not be using counterfeit products."
Meanwhile, US Trade Representative Jamieson Greer told CNBC in a separate interview Tuesday that Washington would "look very carefully at how China is proliferating AI development" and work to ensure American companies can compete on a level playing field.
yckim6452@heraldcorp.com
