The AI investment gap: The US spends 23 times more than China, but the market chooses affordability
The Stanford AI Index 2026 analytical report recorded a massive gap in private investment in artificial intelligence between the US and China. In 2025, American companies invested $285.9 billion in the sector, while China's figure was only $12.4 billion. It would seem that US dominance is indisputable, but the real market dynamics force a reconsideration of these numbers.
Despite a 23-fold financial advantage, it is Chinese AI models that are becoming the cause of new political debates in Washington. This concerns the Kimi K3 model from Moonshot AI, distributed with open weights. Its sudden success in programming tests triggered a collapse in the stocks of American chip manufacturers and revived discussions about imposing strict restrictions.
Last year, the US Department of Commerce considered the option of adding Chinese laboratories to the Entity List — the same trade blacklist that Huawei was placed on in 2019. Alternative ideas included publishing threat warnings from the NSA and discussing the legal liability of American platforms for hosting Chinese neural networks. However, at that time, the line of supporting innovation prevailed, and radical measures were abandoned. Now, hardline advocates have new arguments.
Why Kimi K3 is Causing Panic in Washington
White House AI advisor David Sacks directly pointed to the essence of the conflict: closed market leaders, already generating significant revenue from AI models, are interested in eliminating open-source competitors. Kimi K3 is a prime example. Moonshot suspended new subscriptions just 48 hours after launch due to overwhelming demand, and the company is now preparing for an IPO in Hong Kong.
The key factor in the success of Chinese models is price. DeepSeek V4 Pro charges only $0.87 per 1 million output tokens, while Claude Fable 5 from Anthropic costs $50 for the same volume. The result was immediate: Coinbase CEO Brian Armstrong reported that the exchange's transition to GLM 5.2 and Kimi K2.7 Code halved corporate AI expenses. This is a powerful signal for the entire market.
What the Funding Gap Hides
When analyzing the data, one critical nuance must be considered: official figures do not include direct government injections from Beijing. From 2000 to 2023, Chinese government funds directed approximately $184 billion into the local AI sector. Thus, the real investment gap may be significantly smaller than it appears at first glance.
History is repeating itself. In January 2025, the release of DeepSeek cost Nvidia a record $589 billion in market capitalization in a single day. It was after this collapse that the first discussions of a possible ban emerged. Now, following the launch of Kimi K3, the pressure from restriction advocates has intensified again. However, a complete ban is practically unfeasible: the weight coefficients have already been placed in open repositories and cannot be removed. A ban would only increase costs but would not stop the spread of the technology.
My analysis: The financial gap in favor of the US is obvious, but the decisive factor is not the volume of investment, but the efficiency of the technologies being created. China demonstrates that with a sound strategy and state support, competitive results can be achieved at significantly lower costs. The market is voting for affordability, and this trend will only strengthen, regardless of Washington's regulatory decisions.