Artificial Intelligence ETF Huafu(515980) consolidates and prepares for a rebound. The AI industry is currently in a phase of policy, technology, and demand resonance, and the industry outlook still has room for growth.

In the on-exchange ETF space, as of 13:18 on April 2, 2026, the CSI Artificial Intelligence Industry Index (931071) fell 3.61%. For constituent stocks, Xinghuan Technology led the declines, followed by UCloud, Guanghuan New Network, AOFI Data, and Data Harbor. During trading, Huafu Artificial Intelligence ETF (515980) recorded a turnover rate of 2.16% with trading volume of RMB 191 million. Looking over a longer period, as of April 1, 2026, Huafu Artificial Intelligence ETF has gained 70.02% over the past year.

On the policy front, the Government Work Report has made arrangements for “AI+” for three consecutive years. From the first proposal of the “AI+” Action in 2024, to continued implementation in 2025, and to this year’s clear call to “build a new form of intelligent economy,” seize opportunities for the development of artificial intelligence, and expand both the breadth and depth of AI enabling thousands of industries and fields—so as to open up new space for economic growth as soon as possible, cultivate new business models, and strengthen new momentum.

On industry developments, on April 2, after the Qwen AI glasses were released for the first time, they underwent an OTA upgrade and launched the first batch of “AI office” capabilities. By deeply integrating with Taobao Flash Purchase and Alipay, it supports high-frequency daily-life services such as mobile top-ups, scanning for bike rides, parking fee payments, and voice ordering takeout. On the same day, Alibaba released its next-generation large language model Qwen3.6-Plus. In authoritative evaluations such as the SWE-bench series testing for agentic programming and Claw-Eval for real-world agent tasks, Qwen 3.6’s programming performance exceeds that of GLM-5, Kimi-K2.5 and other models by more than 2 times, even 3 times in terms of parameter count—making it the domestic model with the strongest programming capability at present, approaching the most powerful global programming models like the Claude series.

Worth noting, Huawei senior executives said that AI will be the biggest development opportunity in the next decade, or even longer cycles, and also the most certain strategic opportunity. Dongxing Securities believes that the AI industry is currently in a three-dimensional resonance stage among policy, technology, and demand; domestic chips and leading cloud-computing players are gradually validating their performance, while large enterprises’ ongoing CapEx investment continues to raise the certainty of industry development, leaving further room for improvement in industry sentiment.

Frost & Sullivan forecasts that in 2025, China’s AI智算 GPU market size will reach RMB 171.20 billion. From 2025 to 2029, the compound annual growth rate is as high as 56.7%, reaching RMB 1,033.34 billion by 2029. Driven by the dual forces of increased capital expenditures from cloud providers and faster construction of智算 data centers, the pace of autonomous and controllable AI chip development has significantly accelerated. Domestic vendors are expected to build on breakthroughs in core technologies and deployment across multiple application scenarios to continuously increase their market share.

Huafu Artificial Intelligence ETF (515980) tracks the CSI Artificial Intelligence Industry Index. Leveraging the three-layer closed-loop index construction logic of the underlying index plus the advantage of quarterly rebalancing that stays closely aligned with industry trends, it is a high-quality small-cap index fund for allocating to A-share AI assets. In March, the CSI Artificial Intelligence Industry Index (931071.CSI) completed its latest quarterly regular re-sampling. After this adjustment, the index’s core structure remains highly stable and balanced: the ratio of computing power (hardware foundation) to applications (software scenarios) still stays at the “golden allocation” of 6:4, while the weights of the three key sub-segments—AIGC applications, optical modules, and ASIC chips—still each account for more than 20% in total, together occupying a dominant position in the index. This establishes a foundation under which the index not only keeps up with the certainty of computing infrastructure buildout, but also does not miss the possibility of application surges.

Off-exchange investors without a stock trading account can choose the Huafu Artificial Intelligence ETF Feeder Fund (Class A 008020, Class C 008021).