US efforts to restrict access to advanced AI models may paradoxically accelerate China's rise in artificial intelligence, according to recent analysis from CNBC. As reported by CNBC's Deirdre Bosa, the concern is that policies designed to protect America's lead may instead encourage developers around the world to build on Chinese open-source models. This represents a fundamental shift in the AI competition landscape, where US-China AI model performance gap has effectively closed, with US and Chinese models trading the lead multiple times since early 2025. By March 2026, Anthropic's leading model held a margin of just 2.7% over competitors, as noted in Stanford University's AI Index Report 2026. Bosa frames this as a genuine dilemma, noting that tight controls protect near-term national security interests related to weapons-relevant compute and sensitive model capabilities, but the same controls may push global developers toward Chinese open-source alternatives that the US cannot influence once entrenched.
Chinese AI infrastructure builders are systematically abandoning Nvidia's advanced accelerators in favor of domestic silicon, marking a significant shift in the global AI supply chain. According to a Bloomberg Intelligence survey of 60 executives at Chinese software, finance, manufacturing and retail companies, executives plan to allocate 46% of their AI accelerator budget to domestic products over the next 12 months, up from 30% currently. The survey reveals that 80% of executives say their total infrastructure spending is running over-budget this year, primarily due to high costs of AI-related projects. China's biggest AI infrastructure builders - Tencent Holdings, Alibaba Group, and Huawei Technologies - are best-positioned to capitalize on this shift, with Hygon Information Technology and Cambricon Technologies also being evaluated by respondents. As Bloomberg reports, China is allocating roughly 2 trillion yuan ($294 billion) to build data centers across the country in the next five years, with at least 80% of core technologies, such as chips, to be supplied by domestic companies.
The artificial intelligence industry is experiencing a fundamental split in how companies approach building and commercializing AI systems. According to reports from Business Standard, US-based AI software firm Palantir's chief executive, Alex Karp, recently criticized the token-based pricing model used by OpenAI and Anthropic, arguing that enterprises are being pushed towards expensive AI usage. This debate reflects growing concerns about rapidly rising costs as usage scales, particularly for organizations deploying AI across large teams and high-volume workflows. The monetization divide is becoming increasingly critical as open-source is hard to monetize, which is precisely why state-backed strategies can sustain it longer than venture-funded ones can, as noted by CNBC's Bosa. Bosa warns that if Chinese open-source AI becomes the default foundation like Android, China gains lasting control over the next platform's rules.
China treats open-source AI dominance as a strategic national ambition, while US frontier labs like OpenAI and Anthropic depend on charging for API access and usage. As reported by CNBC, Chinese labs have matched US capabilities in sensitive domains like cybersecurity, with Bosa referencing a bug-finding tool from Chinese company 360 Security as comparable to leading US equivalents. This capability parity in sensitive areas closes gaps many US policymakers assumed export controls would maintain. The systemic risk involves if Chinese open-source models become the default foundation the way Android became the default mobile operating system, China could gain influence over the standards, defaults, and rules of the next AI stack, as well as end users. These models have the potential to build switching costs as developers fine-tune the model family, creating a lasting competitive advantage for China.
Despite China's strategic push toward domestic AI infrastructure, a global memory chip shortage is likely to cap the growth of Chinese AI firms. According to the Bloomberg survey, a global memory chip shortage is likely to cap the growth of Chinese AI firms like Semiconductor Manufacturing International Corp. with ChangXin Memory Technologies Inc. poised to reap the benefits. The bottlenecks are shifting from sheer computing power to securing supply of high-bandwidth memory chips that support rapid data transfer. This supply constraint may limit China's ability to fully capitalize on its domestic AI infrastructure buildout, even as companies like Huawei and Hygon gain market share from Nvidia in the domestic market.
