White House Accuses Chinese AI Firm of Stealing U.S. Technology
A senior White House official has directly accused a Chinese artificial intelligence company of stealing proprietary American technology, raising serious concerns about intellectual property theft and national security.
Michael Kratsios, director of the Office of Science and Technology Policy, stated on July 22 that Moonshot AI covertly distilled technology from Anthropic, a leading U.S. AI firm, to build its Kimi K3 model. He called the action 'large-scale, covert industrial distillation aimed at stealing proprietary U.S. technology' and said it is 'unacceptable.'
What Is Knowledge Distillation and Why Does It Matter?
Knowledge distillation is a machine learning technique where a smaller 'student model' learns from a larger, pre-trained 'teacher model.' While legitimate distillation can foster innovation, Kratsios emphasized that covert, industrial-scale theft of American research undermines fair competition and national security.
According to IBM, the technique is used to transfer learning from a large model to a smaller one. However, in this case, the White House alleges Moonshot used a sophisticated internal platform to conduct large-scale distillation against U.S. models, switching between methods to avoid detection.
How Did Moonshot Access Restricted Technology?
Kratsios also accused Moonshot of acquiring servers equipped with Nvidia's GB300 system and accessing GB300s in Thailand, likely to train its AI models. Washington has restricted Beijing's access to Nvidia's H200 chips, which could be used for military applications.
Access to both models was suspended on June 12 after the government issued an export control directive barring their use by foreign nationals, including noncitizen Anthropic employees.
What Are the Implications for Botswana and Local Industries?
This incident highlights the dangers of unchecked technology transfer and the importance of protecting local industries from unfair competition. For Botswana, it underscores the need for strong intellectual property laws and vigilance against foreign entities that may seek to exploit our markets without contributing to our economy.
It also serves as a reminder that economic sovereignty requires safeguarding our own technological innovations and resisting pressures to adopt foreign models that undermine local businesses.
How Did the AI Community React?
Anthropic's head of public policy, Sarah Heck, thanked Kratsios for speaking out, calling the theft 'illicit, adversarial distillation' that supports adversary military and intelligence capabilities. She described it as 'IP theft and industrial espionage.'
Moonshot released Kimi K3 on July 17, claiming it performed on par with Anthropic's Claude Fable 5 and outperformed OpenAI's GPT-5.6 Sol, while costing far less. However, the company struggled to keep the model running for new users shortly after its launch.
Frequently Asked Questions
What is the main accusation against Moonshot AI?
The White House accuses Moonshot AI of covertly stealing Anthropic's technology through large-scale distillation and accessing restricted Nvidia chips via Thailand-based servers.
Why is this a concern for Botswana?
It highlights the risks of intellectual property theft and the need for strong local protections to ensure fair competition and economic sovereignty.
What actions has the U.S. government taken?
The government issued an export control directive barring use of the models by foreign nationals and has restricted access to Nvidia chips for military applications.