Intel CEO Highlights Rapid Adoption of Chinese AI Models as Lower Costs Strengthen China's Global Competitiveness

Intel Chief Executive Lip-Bu Tan said on Monday that the adoption of Chinese frontier AI models has been remarkably fast, pointing particularly to their lower token costs. Tokens are the fundamental units of information processed by AI systems, and their pricing directly influences how much businesses pay to deploy large language models at scale. Lower inference costs allow developers to integrate AI capabilities into customer service, software development, enterprise automation, and other applications without incurring the same operating expenses associated with more expensive models. Tan's observations highlight an increasingly important dimension of global AI competition: commercial success depends not only on developing the most advanced technology but also on making that technology affordable and accessible to a broad customer base. For companies deploying AI across millions of daily interactions, even modest differences in processing costs can significantly affect the economics of adoption.

China's expanding AI user base provides evidence of this rapid adoption. According to a report published by the China Internet Network Information Center in October 2026, the number of generative AI users in China exceeded 700 million during the first half of the year, representing more than half of the country's internet population. Among surveyed users, 76% reported using generative AI to answer questions, while 47.8% used it for image or video processing, 37.6% for text-related tasks, and 32.5% for workplace activities such as preparing summaries, presentations, and meeting records. Separately, China's National Data Administration reported that average daily token calls had surpassed 140 trillion by late March 2026, compared with approximately 100 billion at the beginning of 2024. Although token volume does not directly translate into revenue, these figures demonstrate the extraordinary expansion of AI computing activity and its increasing integration into everyday digital services and business operations.

Chinese technology companies are also strengthening their positions through competitive model development and open-weight strategies. Alibaba's Qwen series, DeepSeek's reasoning models, and Moonshot AI's Kimi platform have attracted growing attention from developers seeking alternatives to proprietary Western systems. In July 2026, Moonshot introduced Kimi K3, a 2.8 trillion-parameter open-weight model that the company said approached the performance of leading American frontier systems. Open-weight models allow developers greater control over deployment and customization, potentially reducing dependence on a single proprietary provider. Market usage indicators also suggest meaningful international adoption. Research published by RBC Global Asset Management noted that the five most-consumed models on OpenRouter during part of July 2026 were all Chinese, although usage on that platform should not be treated as representative of the entire global AI market. Together, these developments indicate that Chinese developers are competing through a combination of model capability, distribution, flexibility, and pricing rather than relying exclusively on technological performance.

The financial implications extend across the technology industry, particularly cloud services and AI infrastructure. Lower AI processing costs can accelerate enterprise adoption, creating opportunities for companies that supply computing capacity, software platforms, and integrated AI services. Alibaba's fiscal 2026 disclosures illustrate this commercial shift, with external revenue growth in its Cloud Intelligence Group accelerating to 40% in the final quarter and AI-related products accounting for approximately 30% of external cloud revenue. However, declining token prices also create challenges for AI developers, which must generate sufficient revenue to support expensive model training, computing infrastructure, and ongoing research. Rapid usage growth does not automatically translate into sustainable profitability, particularly in a market where aggressive pricing is a major competitive tool. For investors, the distinction between expanding AI consumption and monetizing that consumption remains central to evaluating the sector's long-term financial prospects.

Looking ahead, China's ability to translate widespread AI adoption into sustainable economic growth will depend on continued improvements in computing efficiency, domestic semiconductor capabilities, and enterprise commercialization. Restrictions on advanced chip exports have encouraged Chinese technology companies to pursue alternative hardware solutions and optimize their models for more efficient inference. For semiconductor manufacturers such as Intel, the broader expansion of AI deployment also creates demand for processors and computing infrastructure capable of supporting increasingly complex workloads. At the same time, China's policymakers have emphasized the importance of technological self-reliance, industrial AI deployment, and stronger oversight of emerging technological risks. Tan's assessment therefore reflects a broader transition in the AI industry: the competitive landscape is increasingly being shaped by the ability to deliver advanced intelligence economically and deploy it across real-world applications, rather than by model performance alone.