US Moves to Close Cloud Loophole as China Builds Distillation Pipeline
The Commerce Department's BIS wants licenses for AI chip sales to data centers in five countries. Nvidia shares fell 2.9 percent as a Reuters probe exposed China's distillation industry.
Why it matters
- The BIS interim rule requires US licenses for advanced AI chip sales to data centers in Singapore, Malaysia, Thailand, Indonesia and Brazil.
- Nvidia stock fell as much as 2.9 percent after the announcement; the Philadelphia Semiconductor Index dropped up to 5.5 percent.
- Reuters reviewed more than 300 Chinese papers, patents, job postings and government documents to map China's distillation industry.
- Shanghai Jiao Tong University researchers created 10,750 training examples for Qwen2.5-72B using roughly 5,000 responses from OpenAI's o1 in November 2024.
- Treasury Secretary Scott Bessent is threatening sanctions and Entity List designations over covert distillation attacks.
The US Commerce Department's Bureau of Industry and Security (BIS) has issued an interim rule requiring a government license for the sale of advanced AI chips to data center operators in Singapore, Malaysia, Thailand, Indonesia, and Brazil. The rule, reported by Bloomberg, extends physical export bans to remote cloud access — the legal workaround Chinese companies have used for years to rent Nvidia computing power abroad. Nvidia's stock fell as much as 2.9 percent after the announcement. The Philadelphia Semiconductor Index dropped up to 5.5 percent.
The measures take effect once published in the Federal Register. Applications will be reviewed case by case. Subsidiaries of US-headquartered companies are exempt, as are shipments to data centers running only commercial recommendation systems for online retailers or social media platforms.
For years, the US has tried to cut China off from advanced AI processors through strict export controls. Chinese companies responded by renting the computing power they needed through cloud-based remote access from foreign data centers that were allowed to receive US hardware. The new rule aligns remote-access rules with the existing physical export bans.
Why is the rule landing now?
The political trigger was suspected illegal smuggling of physical chips, not cloud rentals. The announcement comes days after the abrupt resignation of Nvidia CEO Jensen Huang, who admitted to an affair with Alice Huang, a former manager at Singapore-based Megaspeed International. US authorities suspect Alice Huang, who now lives in China, of espionage. They are investigating Megaspeed for possible export-control violations after banned Nvidia chips reportedly turned up in a Chinese data center linked to her. Megaspeed denies any wrongdoing. President Donald Trump said Monday that Jensen Huang did not compromise national security.
The scandal set off an open power struggle in Washington over lax enforcement of chip sanctions. Commerce Secretary Howard Lutnick had delayed a Trump-authorized measure against illegal chip diversions, according to Axios, prompting Treasury Secretary Scott Bessent to publicly accuse him of inaction. The Commerce Department's announcement now states with unusual bluntness that the draft rule was written by, and at the direction of, the Treasury Department and the National Security Council.
What did the Reuters investigation find?
A separate exclusive Reuters investigation shows how Chinese developers cope with the years-long shortage of top-tier domestic hardware: an industrial sector built around model distillation. Reuters reviewed more than 300 Chinese papers, patents, job postings, and government documents, showing for the first time how heavily China relies on distillation as a stopgap. What started years ago as an academic workaround has grown into an industry with a highly developed division of labor.
The technique does not make powerful chips like Nvidia's unnecessary. But it drastically reduces developers' own compute load, because a single expensive query to a US teacher model replaces repeated computation. Chinese firms now build their own software to automate the extraction of answers, code, and problem solutions from leading US models. Specialized workers then process the material to train smaller, far cheaper local student models.
The paper trail runs deep:
- In November 2024, researchers at Shanghai Jiao Tong University created 10,750 training examples for Alibaba's Qwen2.5-72B. They drew on about 5,000 responses from OpenAI's o1 and translations from GPT-4o mini — and warned at the time of a looming dependence on outside innovation.
- In April 2025, Alibaba described its systematic use of GPT-4 and GPT-4o for foreign-language training data.
- In February 2026, Jiao Tong University and food-delivery company Meituan used GPT-4o to train smaller models for coordinating coding agents.
Is Beijing directing this?
Policy directives from Beijing are driving the growth. The government wants to deploy smaller, low-cost AI systems across China's manufacturing sector. Official government documents do not directly instruct anyone to copy American models. Still, an "AI+Manufacturing" package published by eight ministries in January, and initiatives in Shanghai and Shenzhen, explicitly promote distillation and other techniques that reduce computing needs. In practice, this state-driven push for efficiency drives Chinese developers to tap the most powerful US models as a data source at scale.
Ahead of the APEC summit in Shenzhen, the practice is becoming a geopolitical flashpoint. Distillation is not illegal, but it violates the terms of service of US developers like OpenAI and Anthropic. Both companies have long leveled accusations against Chinese AI labs, documenting the massive volumes of extracted data in detailed reports and deploying technical countermeasures. Anthropic recently said it now successfully blocks most such distillation attacks.
How is Washington responding?
US officials confronted Chinese regulators in Chengdu as early as July, according to Reuters, because Washington sees mass distillation as the main reason for China's rapid AI catch-up. Chinese insiders also point to their own advances in algorithms and model design.
Treasury Secretary Bessent is now threatening sanctions and Entity List designations if covert distillation attacks cross the line into intellectual property theft. The threat carries weight beyond chips: Entity List placement can cut a company off from the US financial system and from any supplier using US technology, the same mechanism already used against Huawei.
Chinese developers, uncertain about where the line sits, recently asked Beijing's Ministry of Industry for clear guidelines on using foreign models. The request signals that even Chinese firms cannot predict how enforcement will sharpen — or when the teacher models they depend on will start refusing to answer.
Two fronts are now converging. The BIS rule closes the rental of US silicon abroad; the Treasury threat targets the extraction of US model output itself. If both hold, China's stopgap pipeline loses its two main inputs at once, and the question shifts from whether Beijing can catch up to how much of its catch-up was borrowed.
Original: bis.gov
More from Rebecca Stone
Show full bio
Correspondent covering companies, deals and strategy in artificial intelligence and machine learning at AI In Context.
728 articles
Related articles
- New U.S. Rules Close Cloud Loophole in China AI Chip Export Controls
- White House tightens AI chip exports to China via overseas data centers
- Chinese AI Firms Rush to Lock Up Overseas Nvidia Capacity
- BIS Rule Targets Remote Access to Advanced AI Chips
- Trump Defends China Chip Sales and an Interim Export Rule