US Alleges China's Moonshot Used Anthropic To Build Powerful AI

The US government's Office of Science and Technology Policy (OSTP) Director Michael Kratsios has alleged that China's Moonshot AI used Anthropic's powerful Fable model to develop its K3 model. The K3 model is an open-weight model that has been making waves for both its capabilities and the fact that it is openly available, unlike frontier models from US companies such as Anthropic and OpenAI. Simply put, frontier models are the most advanced AI systems currently available, while open-weight models give developers access to the model's underlying weights, making it possible to customise and run them independently rather than through the company's own service.

"We have information that Moonshot AI distilled Anthropic's Fable for the development of its K3 model," Kratsios wrote on X.

He explained: "To do this they developed a sophisticated internal platform to conduct large scale distillation against U.S. models, allowing them to quickly switch between multiple methods of access to avoid detection."

AI distillation, also known as knowledge distillation, is a technique in which a large, complex and computationally expensive model, known as the "teacher", trains a smaller and faster model.

US President Donald Trump's tech advisor highlighted that legitimate AI distillation used to create smaller, more efficient models is fine and plays a vital role in this open innovation ecosystem. "However, large-scale, covert industrial distillation aimed at stealing proprietary U.S. technology and undermining American research is unacceptable."

Kratsios also alleged that Moonshot AI has acquired servers powered by Nvidia's latest GB300 AI chips and has also accessed GB300 systems in Thailand, likely to train its AI models. The GB300 is Nvidia's newest high-end AI computing platform, designed to train and run advanced AI models at scale.

Moonshot AI has not responded to the accusations made by the United States government.

Leading AI researcher and expert Ethan Mollick, who is also a professor at the Wharton School of the University of Pennsylvania, said, "I have no inside information, but every sign so far is that there is growing tension about open-weight models between the US & China: the US announcing that they reserve the right to act against distilled models, then saying Kimi distills. Contradictory reports from China, etc."

ChrisGPT on X-a popular pseudonymous artificial intelligence commentator and insider on the platform, widely known for frequently sharing early rumours, leaks and insights-commented, "Kimi K3 was already in internal evaluation by April/May, before Fable 5 had been available long enough to be plausibly distilled. I'll wait for the evidence from the USG (United States Government) if they'll publish it."

The Dragon Vs Captain America

A recent study by the UK's AI Security Institute suggests China's leading open-weight AI models are catching up with America's frontier systems faster than before.

According to the research, Chinese models such as Z.ai's GLM-5.2 and DeepSeek V4-Pro now perform at a level comparable to frontier closed models that were released just four to seven months earlier. For most of 2025, the gap was estimated to be around six to ten months.

The findings reflect a broader trend that AI experts have been tracking over the past few months.
"The open-source gap is collapsing exponentially. What took 10 months now takes four, and soon there won't be a gap at all," Srinivas Padmanabhuni, CTO of AiEnsured, told NDTV.

Also read: Chinese AI Is Catching Up To The US Faster Than Before: Why It Matters

Not Everyone In The US Agrees

Even as the Trump administration weighs a tougher stance on Chinese AI, not everyone in Silicon Valley supports restricting access to Chinese open-weight models.

Nearly 200 startup founders and investors have urged the White House not to block access to Chinese open-weight AI models, arguing that many American startups rely on them because they are affordable and widely available. The group instead called for narrowly targeted safeguards that address national security concerns without cutting off access to open AI models already available globally.

The debate highlights a growing divide within the US AI industry. While some frontier AI companies and policymakers have pushed for tighter controls on Chinese AI, many startups argue that broad restrictions could end up hurting American innovation more than China's.

This Dragon Breathes Serious Fire

Kimi K3, developed by Beijing-based startup Moonshot AI, shot to the top of Arena's AI coding leaderboard within hours of its launch, becoming the first Chinese model to claim the No. 1 spot on the widely watched ranking. The leaderboard is widely followed by AI researchers and developers as a measure of how well different AI models perform on coding tasks.

The achievement has reignited debate over whether the US can maintain its lead in AI. For companies like OpenAI and Anthropic, the concern stretches beyond the fact that Chinese models are getting better. Many Chinese models are also significantly cheaper and, in many cases, open-weight. That allows developers and businesses to download, customise and run them on their own infrastructure instead of paying for proprietary cloud-based AI services.

That combination is likely to make them increasingly attractive to enterprises looking to reduce AI costs.

The comparison many in Silicon Valley are making is with DeepSeek. Earlier this year, DeepSeek stunned the AI industry after releasing a highly capable model at a fraction of the cost of its American rivals. The launch rattled investors and briefly wiped hundreds of billions of dollars off the value of major US technology companies.

Some believe Kimi K3 could have a similar impact.

Anastasios Angelopoulos, who runs the Arena benchmark, said the model could force investors to rethink assumptions about the AI industry if businesses increasingly choose free, customisable Chinese models over expensive proprietary ones from companies such as OpenAI and Anthropic.

David Sacks, the White House's AI adviser, has also weighed in, arguing that the latest developments show the US cannot afford to slow down AI development with excessive regulation.

Others see Kimi K3 as evidence that Chinese AI companies are no longer simply copying American systems.
Dean Ball, who recently served as an AI adviser in the White House before joining OpenAI as head of strategic futures, described Kimi as a genuinely capable model.