US accuses Chinese AI firms of 'malicious' copying of AI technology
U.S. agencies say six China-based firms used millions of queries to extract capabilities from frontier models and train their own systems.
- On Tuesday, the National Security Agency, FBI, and Cybersecurity and Infrastructure Security Agency accused China-based AI firms of "industrial-scale" data extraction, claiming they maliciously copied technology from American-made AI models using "distillation."
- The method called "distillation" allows companies to transfer capabilities from larger models to smaller ones to cut training costs, which agencies described as "aggressive, malicious and targeted."
- Agencies identified DeepSeek, Moonshot, Alibaba, MiniMax, and StepFun as companies utilizing American-made AI variants to train their products, an activity occurring "likely with Chinese government awareness."
- These accusations emerge ahead of Chinese President Jinping's upcoming visit to WASHINGTON, creating significant diplomatic friction as tensions escalate.
- US-China relations face further pressure as both nations prepare for a mid-September dialogue concerning AI safety risks, complicating bilateral negotiations.
56 Articles
56 Articles
The U.S. government publicly criticized major Chinese artificial intelligence (AI) companies, stating that they are acquiring technology by “aggressively and maliciously distilling” cutting-edge U.S. models. Ahead of the summit between U.S. President Donald Trump and Chinese President Xi Jinping expected to be held on the 24th, the two sides regarding AI hegemony...
Anthropic, OpenAI, as well as Donald Trump's scientific and technological advisor, Michael Kratsios, have already accused several Chinese AI companies of looting since the beginning of the year. But the document published jointly by the IT Security Agency and Infrastructure (CISA), the NSA Intelligence Agency and the FBI (Federal Police) is by far the most comprehensive and detailed to date. All the important actors of the Chinese AI, from Aliba…
US claims Chinese AI firms are carrying out ‘industrial-scale’ theft of trade secrets - The Tribune
US federal agencies claim Chinese AI firms are committing "industrial-scale" intellectual property theft through unauthorised model distillation using proxy transfer stations. Officials and companies like Anthropic warn of massive covert data extraction, prompting threats of potential US sanctions and increased cybersecurity targeting.
Beijing strongly rejects US allegations that Chinese companies use knowledge distillation to copy US artificial intelligence (AI) products, Liu Chang, spokesperson for the Chinese Embassy in Washington, told TASS
“The US side’s hyping of the so-called ‘distillation’ concept is a deliberate attack on China’s development and progress in the AI industry,” he emphasized, commenting on allegations by US law enforcement and intelligence agencies that...
This website is Cedarnews.net. To learn more about the latest news and technology, visit our site. US administration condemns six Chinese companies for "distilling" AI models; logo of Chinese AI startup Deepseek (AFP-Jiji) [Washington, Jiji Press] The Trump administration on the 8th condemned six Chinese companies for "distilling" models developed by US artificial intelligence (AI) companies, improperly training them. This website is Cedarnews ا…
Coverage Details
Bias Distribution
- 57% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
