US and China vie for global artificial intelligence market share

A new analysis of website traffic for large language model services suggests United States providers still dominate global artificial intelligence use, despite a brief surge from China’s DeepSeek.

The article examines the gap between what new artificial intelligence technologies are capable of in principle and which models are actually being used across the global economy. It discusses research by Austin Horng-En Wang and Kyle Siler-Evans titled “U.S.-China Competition for Artificial Intelligence Markets: Analyzing Global Use Patterns of Large Language Models,” published by the RAND Corporation on January 14, 2026. Their work looks beyond headline-grabbing technical advances to focus on real-world adoption, asking how competition between United States and Chinese providers is playing out through user behavior around the world.

The researchers analyze “website traffic data across 135 countries from April 2024 through May 2025,” tracking “monthly website visits from each country to seven U.S.-based and thirteen China-based LLM service websites.” They acknowledge that these numbers are imperfect, because use of downloaded open-source artificial intelligence tools does not appear in web traffic statistics. The period is notable because in January 2025 a Chinese company called DeepSeek launched an artificial intelligence tool with capabilities that considerably exceeded expectations, which provided an opportunity to see whether a sudden improvement in Chinese technology would quickly shift usage patterns.

Traffic data show that Chinese websites were getting about 2-3% of the total visits in 2024, that with the arrival of DeepSeek the Chinese share rose as high as 13% in February 2025, and that by August 2025 it had sagged back to about 6%. The growth in use of artificial intelligence sites in summer 2025 mostly happened at United States-based websites, suggesting that United States providers maintained a strong lead even after DeepSeek’s launch. The authors explore whether pricing, language support, or diplomatic ties can explain this dominance and find little support for those explanations, noting that while paid subscriptions to United States artificial intelligence services cost more, many users still rely on free access. Given the lack of convincing alternatives, they infer that the capabilities of United States-based tools are currently better, while emphasizing that the DeepSeek episode reveals users around the world are not locked into providers from any given country and are willing to experiment when new options appear.

55

Impact Score

AMD and Rackspace plan dedicated AI compute rollout

AMD and Rackspace have finalized a phased deployment for dedicated AMD-based compute across Rackspace data centers. The capacity is aimed at regulated enterprise workloads, including clinical AI and large-scale inference.

Lexar tests SSD offloading for local AI models

Lexar is developing an AI-focused SSD approach designed to cut DRAM demand when running large language models on consumer PCs. Internal tests show the company’s storage offloading can load models that traditional local frameworks struggle to run with limited memory.

NVIDIA Blackwell leads MLPerf Training 6.0

NVIDIA’s latest MLPerf Training 6.0 results put Blackwell across every benchmark in the suite, including new MoE workloads. Partner systems from Microsoft Azure and CoreWeave highlighted large-cluster runs on Llama 3.1 405B and DeepSeek-V3 671B.

HPE and NVIDIA expand AI Factory for agentic systems

HPE and NVIDIA are adding agent tooling, confidential computing and updated accelerated systems across the HPE AI Factory portfolio. The expansion targets production deployments that need governance, secure data handling and integrated networking.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.