The AI War: China’s Efforts to Bypass US Export Controls and Lead the Pack
The intensifying competition between the United States and China in artificial intelligence continues to reshape global technology supply chains and strategic balances. Recent incidents, including the seizure of advanced AI chips in Malaysia and accusations of large-scale model extraction by Chinese entities, have spotlighted efforts to circumvent export controls and intellectual property protections.
These developments underscore the complex dynamics of the US-China AI race, where hardware restrictions meet sophisticated software techniques. As both nations invest heavily in the sector, questions persist about the effectiveness of controls and the pace of indigenous innovation in China.
Key Takeaways:
- US-China AI competition intensifies, with export controls and smuggling attempts highlighting the complex dynamics of hardware restrictions and sophisticated software techniques.
- China’s indigenous AI capabilities gain momentum, driven by increased research and development spending, open-weight models, and AI deployment across various sectors.
- Smuggling networks adapt to evade export controls, involving relabeling of GPUs and high-level industry figures, despite Beijing’s push for self-reliance.
- Export controls face challenges in enforcement, with third-country transshipments and obfuscation methods exploiting loopholes.
- Intellectual property concerns persist, with distillation attacks and model extraction highlighting the need for enhanced security measures.
Malaysia Chip Seizure Highlights Smuggling Challenges
In June 2026, Malaysian customs officials at Kuala Lumpur International Airport intercepted 72 server units containing advanced AI chips valued at approximately 52.9 million ringgit ($12.93 million). The shipment was falsely declared as ordinary computer components and appeared destined for re-export to another Asian country.
Authorities described the operation as a violation of Malaysia’s Strategic Trade Act. The incident aligns with broader US efforts to curb the flow of restricted Nvidia chips and related hardware to China amid export controls designed to limit advanced AI capabilities.
Reporting indicates that such smuggling networks have involved sophisticated tactics, including relabeling of graphics processing units (GPUs) and involvement of high-level industry figures. For instance, earlier cases have referenced executives linked to major server manufacturers. These operations demonstrate the persistent demand for cutting-edge US technology despite Beijing’s push for self-reliance.
US export controls on advanced semiconductors, updated multiple times since 2022, aim to restrict access to high-performance chips critical for training large AI models. However, enforcement challenges persist through third-country transshipments and obfuscation methods.
Anthropic Accusation Spotlights Model Distillation
In a separate development, Anthropic accused operators affiliated with Alibaba and its Qwen AI division of conducting the largest known distillation attack against its Claude model. The campaign reportedly involved nearly 25,000 fake accounts and over 28.8 million interactions between April 22 and June 5, 2026.
Distillation involves querying a leading model extensively to extract capabilities and transfer them into a rival system, often at significantly lower cost than independent training. Anthropic described the activity as “brazen” and “illicit,” targeting strengths in areas such as software engineering and agentic reasoning.
Alibaba has denied wrongdoing in public statements. The episode reflects longstanding concerns over intellectual property in technology transfers between the US and China, a pattern dating back decades across various sectors.
Experts note that while countermeasures like data poisoning can raise the cost of such attacks, the efficiency gains for the extractor remain substantial. This technique allows bypassing some data acquisition hurdles inherent in training frontier models from scratch.
China’s Push for Indigenous AI Capabilities
China has significantly increased its research and development spending. In 2025, total R&D investment reached approximately 3.92 trillion yuan (about $569 billion), equating to 2.8 percent of GDP, with plans for continued annual growth.
This investment supports efforts to indigenize semiconductor production and AI model development. Chinese firms have released competitive open-weight models, and the country continues to embed AI across sectors including healthcare, transportation, and manufacturing.
Government policy emphasizes both commercial competitiveness and strategic applications. Initiatives tie into broader economic strategies, aiming to reduce dependency on foreign technology while expanding influence in emerging markets through tech exports.
Despite progress, analysts highlight gaps in cutting-edge fabrication and the continued appeal of advanced US chips, evidenced by smuggling attempts. Beijing has, at times, restricted imports of certain Nvidia products to prioritize domestic alternatives.
US Export Controls and Global Implications
The United States maintains a framework of export controls administered by the Bureau of Industry and Security. These measures target advanced computing commodities, with licensing policies that often presume denial for destinations in China.
Recent adjustments, including proposals for location-tracking features in chips and enhanced third-country oversight, seek to close loopholes. However, the effectiveness of these tools remains under scrutiny as networks adapt.
The competition extends beyond hardware. Discussions in policy circles address open-source strategies, local deployment of models, and the economics of inference versus training. Emerging markets show varying levels of adoption, influenced by cost, accessibility, and geopolitical alignments.
Forex Trading Basics from Fortune Prime Global provides essential insights into trading fundamentals for those navigating related market dynamics influenced by technology shifts.
As a reputable Forex Broker, Fortune Prime Global benefits its clients with reliable access to global markets amid these technological and economic developments.
Broader Context of the AI Race
The US leads in private AI investment, with figures reaching $285.9 billion in 2025 compared to lower reported private totals in China. However, China’s state-supported ecosystem and total R&D outlays present a different picture of overall commitment.
Panelists in recent analyses, including former Pentagon officials and China experts, describe a multifaceted environment involving state, commercial, and other actors. Motivations range from military applications to economic path dependency in global supply chains.
Open-weight models and local AI deployment introduce additional variables, potentially accelerating adoption in regions with infrastructure constraints while raising security considerations. Energy costs and scalability remain key factors in deployment strategies.
Historical parallels exist in previous technology competitions, where espionage, subsidies, and innovation races coexisted. Current dynamics include efforts to build ecosystems around models, influencing not just performance but market penetration and revenue models.
Outlook for AI Development
The incidents in Malaysia and involving Anthropic illustrate ongoing frictions in the US-China AI relationship. While controls aim to preserve advantages, the diffusion of technology through various channels continues.
China’s sustained investment and policy focus suggest a long-term commitment to technological parity or leadership in specific applications. The United States emphasizes protection of foundational advantages alongside domestic advancement.
As the technology evolves, international norms around intellectual property, export compliance, and responsible development may gain prominence. Outcomes will likely depend on the interplay of innovation, enforcement, and global market forces rather than any single tactic.
This episode in the AI race highlights the strategic importance of the sector for economic and security considerations worldwide. Continued monitoring of policy, investment, and technological milestones will shape future developments.