Artificial intelligence Photo: VCG
China has seen a wave of concentrated issuances of artificial intelligence (AI)-themed exchange-traded funds (ETFs) against the backdrop of the country’s fast-advancing technology sector.
Several AI-themed ETFs, including those from Huatai-Pinebridge and E Fund, will open for subscription on Friday. On November 21, regulators approved the issuance of the first batch of seven STAR-ChiNext AI ETFs, the Shanghai Securities News reported on Tuesday. These funds track the CSI STAR&CHINEXT Artificial Intelligence Index (STAR&CHINEXTAI), launched on May 14, the first AI-themed benchmark to span both boards.
Meanwhile, AI Singapore (AISG) is undergoing a major strategic shift, moving away from Meta’s model and adopting Alibaba’s open-source Qwen architecture for its latest Southeast Asian language model project. The decision marks a notable expansion of China’s open-source AI footprint on the global stage, according to a statement, citing media reports, Alibaba sent to the Global Times.
AISG’s newly released “Qwen-SEA-LION-v4” model, announced on Tuesday local time, quickly rose to the top of an open-source leaderboard measuring Southeast Asian language capability. The shift addresses a long-standing bottleneck: earlier open-source models, including Meta’s Llama series, performed poorly in languages such as Indonesian, Thai and Malay, limiting the development and effectiveness of localized AI applications, said the statement.
Analysts said that Singapore’s choice of Alibaba’s open-source Qwen model is closely tied to linguistic fit. With a significant portion of Chinese-speaking population, Singapore benefits from Qwen’s stronger alignment with local data. This gives the model a natural advantage in training quality and language accuracy, making it smoother and more effective for Southeast Asian cases than Meta’s model.
Notably, the shift comes at a time when China’s open-source AI models are rapidly expanding their global footprint and gaining strong market momentum. Chinese-developed AI is also making a fast push into the AI-to-consumer (A-to-C) segment, posting early but striking results.
Alibaba’s Qwen AI assistant app logged more than 10 million downloads in its first week of public testing, surpassing ChatGPT, Sora and DeepSeek to become the fastest-growing AI application to date, China Star Market reported.
Ant Group’s Lingguang app also saw rapid uptake, exceeding 200,000 downloads on day one, 500,000 within two days, and 1 million within four days, topping the App Store’s free utilities chart in China and breaking into the platform’s overall top six. Its popular “Flash Apps” feature briefly went offline due to surging traffic.
As Chinese large-language models accelerate their iteration and AI applications move more quickly from enterprise to consumer use, the industry is entering a new window of A-to-C competition. With demand rising sharply for localized, scenario-based and low-cost AI services, domestically developed AI applications are racing to secure the next major consumer gateway.
Foundation models such as DeepSeek and Qwen are enabling a rapid rise of new AI service hubs including Yuanbao, Doubao and Qwen’s own apps. Backed by nationwide data and application-scenario strategies, as well as recent State Council policies, China is leveraging its market of 1.4 billion people and vast industrial landscape to accelerate model testing, enrich training data and break through the “data wall” ahead of the US, Tian Feng, president of the Fast Think Institute and former dean of Chinese AI software giant SenseTime’s Intelligence Industry Research Institute, told the Global Times.
Domestic chipmakers from Huawei Ascend to Cambricon and SMIC are pushing through key technological bottlenecks and steadily narrowing the gap with the US chips, Tian said.
Bi Qi, chief scientist at China Telecom and a fellow of Bell Labs, told the Global Times that US companies often pursue “ultimate performance” in developing large AI models, which drives costs sharply higher. By contrast, Chinese models, with their strong price-performance ratio, simpler architectures and lower power consumption, are better suited to engineering and commercial needs. In many Global South markets, including parts of Africa, users place greater emphasis on affordability and once basic quality is met, on achieving a viable commercial loop and deeper local cooperation.
Ma Jihua, a veteran industry expert, told the Global Times that the advantages of China’s large AI models — not only their strong uptake across the Global South but also their growing appeal among companies in advanced economies — are strengthening China’s competitive position in the global AI race. From this perspective, Chinese models may gain an even clearer edge over their US counterparts in the coming years.
Ma noted that these dynamics are also fueling the growing popularity of China’s AI-focused investment funds and AI-themed ETFs, as investors look to tap into the country’s rapidly expanding AI ecosystem. New products continue to enter the market amid rising demand for AI-related exposure, he added.