China's Open-Source AI Dominance Threatens US AI Lead Despite Chip Export Restrictions

China's Dominance in Open-Source AI Challenges US Leadership

China open-source AI is building a self-reinforcing competitive advantage that threatens the US AI lead, according to a report published Monday by the U.S.-China Economic and Security Review Commission 1

1

. The U.S. congressional advisory body warns that Chinese large language models from firms including Alibaba, Moonshot, and MiniMax now dominate worldwide usage rankings on platforms like HuggingFace and OpenRouter, driven primarily by their cheaper cost 1

1

.

Despite export restrictions on advanced AI chips imposed by U.S. lawmakers since 2022, Chinese labs have narrowed performance gaps with top Western large language models. The report emphasizes that this open ecosystem enables China to innovate close to the frontier despite significant compute constraints 2

2

. The threat to U.S. AI leadership has intensified as model proliferation creates alternative pathways to AI dominance.

Chinese Models Gain Ground Among US Startups

The scale of adoption reveals the depth of the challenge facing American companies. Around 80% of U.S. AI startups now use Chinese open-source AI models, according to estimates cited in the report 1

1

. DeepSeek's R1 model launched last year quickly overtook ChatGPT as the most downloaded model on the U.S. App Store, while Alibaba's Qwen family of models has surpassed Meta's Llama in global cumulative downloads on HuggingFace 2

2

.

The cost advantages of Chinese models have proven irresistible even to major Western corporations. Siemens CEO Roland Busch stated Monday that there were "no disadvantages" to using Chinese open-source AI to train the German company's AI models specialized for industrial automation, citing their cost advantage and ease of customizing parameters 1

1

. This comes despite warnings from Western research organizations about potential security concerns and political bias toward Chinese government positions in these models.

Shift to Embodied AI Favors China's Data Collection Efforts

As AI's frontiers move from large language models toward agentic AI and the shift to embodied AI, China may be better positioned to capitalize on its mass data collection efforts 2

2

. Beijing's push to deploy AI throughout manufacturing, factories, logistics networks, and robotics generates real-world data that feeds back into model improvement, creating a feedback loop that compounds over time.

"There's a bit of a deployment gap in the embodied AI space between the U.S. and China. That's something that over time compounds itself... We're starting to see that compounding now," Michael Kuiken, the commission's vice-chair, told Reuters 1

1

. Beijing has designated embodied AI as a core future strategic industry, with many leading Chinese humanoid robots firms planning public listings this year. The commission is also monitoring how China uses AI in sectors like biotech, quantum computing, and advanced materials.

While U.S. companies including OpenAI and Anthropic have invested billions of dollars to remain at the forefront, Washington approved exports of Nvidia's second-most advanced chip only in December. The report suggests that export restrictions on advanced AI chips may not be sufficient to maintain American technological superiority as model proliferation and compute constraints force innovation through alternative pathways 1

1

.