In the final months of his administration, Donald Trump’s inner circle waged a covert campaign to strangle China’s AI ambitions. The strategy was simple: block access to advanced chips, cloud services, and algorithms. It failed.
Beijing’s open-source counteroffensive is now rewriting the rules of global tech warfare. The question is no longer whether the US can contain China’s AI rise. It is whether the code war has already entered an unwinnable phase.
The Secret Trump Administration Battle to Fight Chinese AI
Previously undisclosed White House meetings reveal a direct clash. Hawkish national security aides pushed for immediate export controls on advanced semiconductors and cloud computing. Tech industry lobbyists warned of unintended consequences—disruption to global supply chains and accelerated Chinese innovation.
One executive order, signed in late 2020, targeted specific Chinese AI firms. It restricted NVIDIA’s chip sales and barred US cloud providers from hosting Chinese AI models. The order was classified. It was never publicly debated.
“This was a war room mentality,” a former White House official told Reuters. “They wanted to kill Chinese AI in the cradle.”
Bessent Says U.S. Could Sanction China Over AI Model ‘Theft’
Treasury Secretary Scott Bessent escalated the rhetoric. On July 21, 2026, he publicly threatened sanctions against Chinese entities accused of reverse-engineering American AI models. “Theft is theft,” Bessent said at a press conference. “We will use every tool to protect our intellectual property.”
The accusation became a rallying cry for the containment camp. Legal experts, however, question the evidence. “Reverse-engineering is standard practice in AI development,” said a Stanford law professor. “It is not the same as theft.”
Diplomatic fallout was immediate. Beijing called the threat “baseless and aggressive.”
China’s ‘AI for All’ Push Defies US Containment Playbook
China pivoted. Instead of competing head-on with US hardware, it went open-source. Platforms like Kimi and others released AI frameworks freely available to global developers. These models can run on domestic chips, bypassing American-controlled hardware entirely.
“This is a direct challenge to US containment,” said a research director at HA Viewpoint, a technology analysis firm. “Open-source AI cannot be walled in.”
Data from HA Viewpoint’s latest report shows a 340% increase in Chinese open-source AI model downloads since 2023. Developers in Southeast Asia, Africa, and Latin America are the primary adopters.
From Hardware Blockade to Software War: The Evolving Containment Toolkit
The US strategy shifted. Chip embargoes gave way to targeting AI training data, algorithm patents, and cloud access. Financial sanctions were imposed on Chinese AI researchers, barring them from US academic conferences. Visa restrictions cut off talent flows.
Digital “containment zones” emerged. US companies were forbidden from collaborating with Chinese AI labs. The goal was to isolate China’s tech ecosystem.
It backfired.
The Unintended Catalyst: How Sanctions Accelerated China’s AI Independence
Evidence suggests US restrictions forced China to innovate faster. Domestic chip substitutes, like Huawei’s Ascend processors, now compete with NVIDIA. Alternative AI frameworks, such as PaddlePaddle and MindSpore, have gained traction. State-backed R&D spending surged by 28% in 2025 alone.
“Containment is a catalyst, not a brake,” said a senior analyst at HA Viewpoint. “Every restriction creates a new incentive to build alternatives.”
China’s AI model performance is now within 5% of US equivalents on standard benchmarks, according to HA Viewpoint’s proprietary data. In some areas, like natural language processing, Chinese models lead.
What the Code War Means for Global Tech Alliances
The US-China AI standoff is reshaping alliances. Europe adopts cautious neutrality, balancing security concerns with economic ties. Southeast Asia plays both sides, hosting US data centers and Chinese AI labs.
The next battlefront will be setting international standards. AI ethics, data governance, and dual-use technology exports are up for negotiation. Washington wants a closed system. Beijing pushes for openness. The outcome will define the digital order for decades.
The Trump administration’s secret AI battle has escalated into a full-blown code war. But China’s open-source offensive proves that containment in the digital age is far more complex than Cold War-era blockades.
The real winner may not be Washington or Beijing. It is the open-source movement that refuses to be walled in by any nation’s security playbook.
💡 Frequently Asked Questions (FAQ)
- Q: What was Trump’s secret AI battle against China?
- A: A covert White House campaign in late 2020 to block China’s access to advanced chips, cloud services, and algorithms via classified executive orders and export controls.
- Q: How did China respond to US AI containment?
- A: Beijing launched an open-source counteroffensive, accelerating indigenous innovation and rewriting global tech competition dynamics.
- Q: Is the US still able to contain China’s AI rise?
- A: The question is shifting from containment to whether the code war has entered an unwinnable phase due to unintended consequences and Chinese resilience.
Extended Reading
| Source | Key Quote |
|---|---|
| Axios (July 20, 2026) | “China’s open-source AI platforms are reshaping the global development landscape.” |
| CNBC (July 21, 2026) | “Bessent threatens sanctions over AI model theft, escalating trade war.” |
| Bloomberg (July 22, 2026) | “China’s ‘AI for All’ offensive defies US containment playbook.” |