Brokerage firms' annual reports secretly reveal a major competition in financial technology, with leading firms heavily investing in AI, while smaller brokerages are still in the early stages.

Ask AI · How Do Small Brokerage Firms Respond to the Advantages of Leading Institutions in the AI Race?

Interface News reporter | Sun Yizhen

The 2025 annual report provides a clear window for external observers to view the layout of China’s securities industry’s fintech.

As AI technology fully penetrates, “Artificial Intelligence,” “AI,” and “Large Models” have become “high-frequency terms” appearing in brokerage annual reports, widely used across business areas such as investment research and analysis, customer service, and risk management.

Faced with deep internet penetration and a new round of technological revolution driven by large models, securities firms have stepped up their investment in information technology, using digital technology to improve customer experience, optimize business processes, and promote a transition from traditional fee-based models to diversified directions that emphasize professional services, deepen customer relationships, and support networked operations. However, small brokerages still face issues such as the high cost of building large models and the lack of application scenarios that have already been launched, reflecting an increasingly intensified pattern of divergence in the securities industry’s fintech competition.

From the overall investment perspective, the average information-technology investment in the securities industry continues to grow. According to incomplete statistics by Interface News reporters, in 2025, two leading brokerages’ information technology capital expenditures each exceeded 2 billion yuan.

Specifically, after the merger, Guotai Haitong (601211.SH) stands out, with total information technology investment reaching 3.235 billion yuan in 2025; Huatai Securities (601688.SH) recorded information technology investment of 2.679 billion yuan in 2025, up 9.43% year-on-year from 2.448 billion yuan in the previous year.

In addition, according to Interface News statistics, among the brokerages that disclosed detailed information technology investment amounts, at least 7 other brokerages had information technology investment exceeding 1 billion yuan, generally among large and mid-sized brokerages; among them, multiple brokerages saw double-digit growth in this item compared with the same period last year.

Specifically, China Merchants Securities (600999.SH) invested 1.908 billion yuan, with a year-on-year increase of 20.58%, ranking among the industry’s top performers; CICC (601995.SH) had information technology investment of 1.651 billion yuan in 2025, up 4.16% from 1.585 billion yuan in the same period last year; Zhongtai Securities (600918.SH) invested 1.003 billion yuan, accounting for 11.19% of the parent company’s operating income; CCBIB (601066.SH) invested 1.694 billion yuan, an increase of about 11%; GF Securities (000776.SZ) invested 1.548 billion yuan, an increase of 12.41%; Galaxy Securities (601881.SH) invested 1.345 billion yuan, an increase of 7.26%; Shenwan Hongyuan (000166.SZ) invested 1.107 billion yuan, basically flat year over year.

A signal worth focusing on is that Huatai Securities (601688.SH) and CICC (601995.SH) saw their information technology investment decline for two consecutive years in 2023 and 2024, but in 2025 the uptrend returned.

Looking at year-on-year changes, some small and mid-sized brokerages’ IT spending also grew. For example, Hu’an Securities (600909.SH) invested nearly 240 million yuan in information technology in 2024; in 2025, that figure rose to 263 million yuan, accounting for 6.8% of operating income; Everbright Securities (601788.SH) achieved “three consecutive increases,” rising from 545 million yuan in 2023 to 659 million yuan in 2024, and then to 721 million yuan in 2025.

According to the Securities Industry Association’s 《Three-Year Plan for Enhancing Network and Information Security Capabilities of Securities Firms (2023—2025)》, brokerages’ average information technology investment amount must be no less than 10% of their average net profit from 2023 to 2025, or 7% of their average operating revenue. This hard constraint, combined with industry-wide consensus on digital transformation, jointly forms the underlying driving force behind brokerage fintech investment in 2025.

At the same time, as AI technology develops rapidly, technologies represented by large models are accelerating implementation in areas such as wealth management, investment research, and investment banking; one major highlight for the industry is the technology of intelligent agents spreading across multiple business lines and achieving large-scale application.

“Our company’s AI applications have been carried out across seven major business lines, including institutions, investment banking, and operations,” a relevant person in charge at Zhongtai Securities (600918.SH) said in an interview with Interface News. “For example, in investment banking, intelligent writing supports more than 100 types of investment-banking documents, with an adoption rate exceeding 95%; the centralized operations ‘Shan Lu’ model shortens the time required for account opening by more than 40%; the macro risk analysis assistant shortens the time spent processing information by 90%, enabling AI to shift from office-efficiency improvements to innovative empowerment for core-business revenue generation and risk prevention and control. In the wealth management segment, AI marketing, investment advisory, and customer service core scenarios have been implemented.”

In wealth management, Huatai Securities (601688.SH) launched the “AI Zangle” APP focused on trading scenarios; China Merchants Securities (600999.SH) launched a marketing service assistant called “Zhao Xiao Gu”; CCBIB (601066.SH) independently developed the “Octopus” fixed-income client-facing business integration big-data intelligent platform; Galaxy Securities (601881.SH), in its annual report, also mentioned that the company has built a full-scenario digital human service system, serving more than 1.3 million customers in total across the year.

Focusing on the investment research area, China Merchants Securities (600999.SH) noted in its annual report that the company is committed to building an “AI Research Institute,” exploring the application of generative AI in scenarios such as financial report interpretation, report writing, and analyst-assistant research processes; Eastmoney (300059.SZ) launched the “AI Researcher,” which can independently complete research processes such as information collection and analysis; CCBIB (601066.SH) built an integrated intelligent investment research platform of “Smart Research, Smart Q&A, and Smart Data” by leveraging large-model technology.

AI is not only improving the basic work efficiency of investment banking staff—applications on the investment-banking business front are also accelerating deployment. Guotai Haitong (601211.SH) mentioned in its annual report that the company is systematically advancing the construction of a unified investment-banking AI platform, “Zhi Jian,” and is taking a forward-looking layout for a full-process digital and intelligent transformation of investment banking driven by large models; CICC’s (601995.SH) “Zhi Yue” large model has been applied in investment-banking risk control review scenarios.

However, in sharp contrast with leading brokerages, some small and mid-sized brokerages are still at the initial stage in their AI layout.

“Large-model construction requires a lot of investment. The company isn’t as willing as leading institutions to spend money on this,” an IT-related person from a certain small brokerage told Interface News in an interview. At present, the company’s exploration in the field of artificial intelligence is still in an early stage of internal incubation.

According to the person, from the perspective of planning and reporting, at the early stage the company’s AI technology is mainly planned to be used by integrating with middle- and back-office departments—for example, assisting compliance and risk-control reporting analysis, document review, and code generation in the technology line; at the same time, the company is also trying to combine AI with investment strategies for scenarios such as stock-selection analysis.

The contrast between small brokerages’ “ALL IN AI” investment and the rich scenarios already put into practice by leading brokerages also reflects an increasingly intensified divergence in the securities industry’s fintech competition.

“Right now, we haven’t truly built anything, and even basic infrastructure such as graphics cards and servers hasn’t been procured or deployed yet,” the person further said when discussing the progress of large-model deployment. “To see obvious effects, at least until the second half of this year. Then we’ll observe which links can significantly improve efficiency.”

He admitted that, compared with leading brokerages that already have proven deployed cases, his brokerage is one of the later starters and has not yet effectively combined AI with specific business scenarios. “In the future, the company may consider building in-house, and we also don’t rule out adopting a leasing model.”