Chinese AI Industry After Export Restrictions: Competition Shifts to Cost and Scale

Report: China Leads Among BRICS in R&D Expenditure

+

Chinese company SenseTime, restricted by U.S. sanctions, has publicly stated that its focus lies in reducing the cost of multimodal models to maintain competitiveness and expand into markets outside the United States. This reflects a broader trend across China’s high-tech industries: competition is no longer solely about quality, but also about access costs, ease of user acquisition, and the ability to deploy models across diverse hardware platforms. U.S. export and investment restrictions are pushing Chinese firms to enhance efficiency and strengthen technological self-reliance.

• SenseTime was listed by the U.S. Department of the Treasury on the restricted investment list due to alleged involvement in surveillance technologies; the company has publicly denied these allegations.

• In late April 2026, SenseTime announced the release and open-sourcing of its daily-updated SenseNova U1 series, positioning it as a “unified” multimodal solution capable of handling both language and image processing within a single architecture. This approach improves performance while lowering operational costs.

• The economic benefits of this strategy became clearer in the company’s financial results for 2025. SenseTime reported reduced net losses in the second half of the year and positive EBITDA. Given the high cost of computing power, this is crucial for assessing the viability of monetizing artificial intelligence.

• U.S. supply restrictions on advanced computing and semiconductor technologies have made acquiring such components increasingly difficult. Exporting these components now requires individual licenses, and the U.S. Department of Commerce’s Bureau of Industry and Security revised relevant regulations in 2026. As a result, solutions that can be deployed domestically using affordable, non-dependent hardware have become especially critical.

• The intrinsic logic of the Chinese market is reinforcing this shift. Amid model saturation and intense price competition, companies that can rapidly integrate AI into products, cloud services, and enterprise solutions—and scale them to vast user bases—are winning market share. Low model access costs are becoming a key pathway for enterprises to capture market share and enter regions where reliability and affordable service pricing matter more than sheer quality.

The “performance meets requirements” strategy combined with cost-reduction measures enables Chinese firms to expand beyond U.S. markets despite sanction pressures. All these factors are contributing to the formation of an independent Chinese AI ecosystem: even as access to U.S. chips narrows, Chinese companies are compensating through model optimization, architectural innovations, and leveraging large-scale platforms capable of rapidly turning technology into mass production.

+

A report from Russia's Government Financial University shows that during the period 2020–2024, China ranked first among BRICS countries in R&D expenditure.

The report indicates that, according to calculations, China’s R&D spending accounted for 2.45% of GDP between 2020 and 2024. Brazil (1.15%) and Russia (1.02%) ranked second and third respectively among BRICS nations.

Meanwhile, China’s R&D expenditure increased by 1.49% in 2024 compared to 2020. For comparison, other BRICS countries saw the following growth during the same period: Brazil grew by 0.14%, Iran and Indonesia by 0.21%, while others declined—Russia fell by 0.16%, India and South Africa by 0.09%.

Original source: toutiao.com/article/1864570683088908/

Disclaimer: The views expressed in this article are those of the author(s) alone.