WorldPolitics
US Accuses Chinese AI Firms of Systematic Theft From American Rivals
The United States has accused several Chinese artificial intelligence companies, including DeepSeek and Moonshot AI, of systematically extracting proprietary capabilities from U.S. AI models. The advisory, issued by a U.S. cyber defense agency, names Alibaba, MiniMax, StepFun, and Z.AI as allegedly using millions of interactions to copy features from models by OpenAI, Anthropic, Google, and xAI.
Washington has accused leading Chinese artificial intelligence labs of systematically stealing capabilities from American AI models, according to an advisory from a U.S. cyber defense agency. The advisory names DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun, and Z.AI, saying they used millions of exchanges and requests to extract proprietary functions from U.S. models such as those developed by OpenAI, Anthropic, Google, and xAI.
The statement said "China-based artificial intelligence (AI) companies are conducting systematic extraction of proprietary functionalities and capabilities of U.S. AI companies' models." It described a process known as distillation, where an AI model is trained to replicate another model's output.
The accusation comes ahead of an expected meeting this month between U.S. President Donald Trump and Chinese President Xi Jinping at the White House, where AI could be a topic of discussion. This is not the first time the U.S. government and American AI firms have accused Chinese companies of such theft, as similar allegations have been made previously.
The advisory did not provide evidence publicly, and the accused companies have not yet publicly responded to the specific claims, according to the report.
- U.S
- OpenAI
- DeepSeek
- Chinese AI
From
WorldPolitics
US Accuses Chinese AI Firms of Systematic Theft From American Rivals
The United States has accused several Chinese artificial intelligence companies, including DeepSeek and Moonshot AI, of systematically extracting proprietary capabilities from U.S. AI models. The advisory, issued by a U.S. cyber defense agency, names Alibaba, MiniMax, StepFun, and Z.AI as allegedly using millions of interactions to copy features from models by OpenAI, Anthropic, Google, and xAI.
Washington has accused leading Chinese artificial intelligence labs of systematically stealing capabilities from American AI models, according to an advisory from a U.S. cyber defense agency. The advisory names DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun, and Z.AI, saying they used millions of exchanges and requests to extract proprietary functions from U.S. models such as those developed by OpenAI, Anthropic, Google, and xAI.
The statement said "China-based artificial intelligence (AI) companies are conducting systematic extraction of proprietary functionalities and capabilities of U.S. AI companies' models." It described a process known as distillation, where an AI model is trained to replicate another model's output.
The accusation comes ahead of an expected meeting this month between U.S. President Donald Trump and Chinese President Xi Jinping at the White House, where AI could be a topic of discussion. This is not the first time the U.S. government and American AI firms have accused Chinese companies of such theft, as similar allegations have been made previously.
The advisory did not provide evidence publicly, and the accused companies have not yet publicly responded to the specific claims, according to the report.
- U.S
- OpenAI
- DeepSeek
- Chinese AI
From