Earnings calls flooded with AI! Banks collectively betting on it, with security as the opening move

How are banks responding to the data privacy challenges in AI applications?

Loan services that once required hours of waiting in line and filling out complicated forms can now be precisely pushed out with limits and interest rates within minutes through a mobile banking app; at bank branches, tellers are engaging in real-time conversations with AI assistants to solve business problems, making “silicon-based colleagues” an indispensable working partner for frontline employees. From state-owned banks to local banks, and from top-level strategic planning to frontline implementation, AI is penetrating all aspects, driving banks from “human governance” to “technological governance.” In recent earnings conferences held by major banks, executives frequently mentioned AI strategies, marking a comprehensive industry transformation led by AI. This transformation not only reshapes the business models and collaboration methods of banks but also pushes financial services back to their essence, extending service reach infinitely and continuously refining service granularity. However, alongside the opportunities, challenges persist, as issues such as data silos, privacy protection, and regulatory adaptation urgently need to be addressed.

From “tool” to “silicon-based colleague”

Against the backdrop of increasing complexity in banking operations, AI assistants have become crucial in breaking down “knowledge silos” and enhancing frontline service capabilities. At the 2025 earnings conference, China Construction Bank’s Vice President Lei Ming revealed a set of data: by the end of 2025, the AI assistant’s coverage rate in responding to inquiries at China Construction Bank branches had reached 99.42%, with an average daily access volume exceeding 100,000 visits.

This means that when employees encounter business challenges and seek help from the headquarters or management departments, in most cases, artificial intelligence will provide solutions first. This “super tutor” with the “strongest brain” is always on standby, patient, and available year-round, transforming the collaboration model within banks.

The transformation is particularly evident in Industrial and Commercial Bank of China’s “ICBC Smart Surge” model technology system, where the bank has scaled over 500 AI applications across more than 30 business areas, with AI digital employees carrying a workload equivalent to 55,000 people annually. These “employees” who do not require salaries can operate “24/7,” relieving significant business pressure. China Merchants Bank leverages large model technology to enhance the efficiency of financial reimbursements, processing 1.4085 million paperless reimbursement requests by the end of last year, a year-on-year increase of 23.76%. Industrial Bank’s AI programming assistant covers 90% of R&D personnel; the Morning and Evening Meeting Intelligent Assistant covers over 1,500 departments and institutional branches.

AI has become an indispensable “working partner” for employees across all lines in banks. China Merchants Bank has developed a retail series of “small assistants” in the retail line, continuously empowering customer managers and middle office teams in client management, operational analysis, and wealth research scenarios. In the wholesale line, they created “CRM small assistants” to help corporate customer managers improve service quality. The risk line developed “Risk Small Assistants,” embedded in operational processes to achieve intelligent risk management. The operations line created “Operational Small Assistants,” implementing applications for digital assistants, operational knowledge Q&A, intelligent business review, intelligent service practice, and intelligent risk event analysis. By the end of 2025, the user coverage rates for corporate customer managers, loan personnel, and operational personnel using their respective small assistants reached 80.13%, 80.32%, and 100%.

Industrial Bank is also ramping up its efforts, building over 160 AI open capabilities, launching over 260 AI application scenarios, and achieving 24/7 intelligent interaction through AI digital customer service across 13 channels, serving over 55 million customers; AI marketing strategies have reached over 21.39 million customers cumulatively.

On the frontline battlefield directly facing customers, AI is redefining the boundaries of “service.” Traditional banking services are often limited by labor costs and can struggle to achieve true personalization. However, Bank of Communications has added AI product interpretation and AI-assisted research generation functions to its wealth management system, meeting the personalized asset allocation needs of a wide range of customers; Ping An Bank has upgraded its “AI + T + Offline” service model, enhancing the application of digital tools like AI assistants and intelligent voice calls to improve remote banking service efficiency; CITIC Bank utilizes small model + large model capabilities to empower corporate account openings and changes, fully establishing a new operational model with more than double the efficiency in business centralization.

At the 2025 earnings briefing, Industrial Bank Chairman Lü Jiajin highlighted future trends, viewing the AI era as one where “silicon-based life will largely replace carbon-based life in work. By feeding some AI entities financial knowledge related to funds, retail, and interbank, one person can play multiple roles. In the future, customer managers will no longer distinguish between corporate, retail, and interbank types.”

AI’s “full-scale penetration” into the industry

The core logic of this AI “penetration war” is the transition of banks from traditional “human governance” to efficient “technological governance.”

From the “aircraft carrier-level” turnaround of state-owned banks to the agile breakthrough of joint-stock banks, and to the precise positioning of local banks, AI is no longer just an embellishment but has become the nervous system permeating the capillaries of business operations.

In top-level design, major banks have begun to take action. According to the latest data disclosed in annual reports, Industrial and Commercial Bank of China is implementing the “Leading AI + Action” plan at the group level in 2025, empowering four core scenarios: investment trading, marketing expansion, risk control, and operational efficiency; Postal Savings Bank has opened 10 major categories of 24 general AI capabilities to its branches, forming the “AI2ALL” digital ecosystem that achieves “external full reach + internal efficiency for all.”

China Merchants Bank has proposed the “AI First” concept, placing AI in a position of “priority, leadership, and initiative” on the bank’s strategic chessboard. The change in top-level design also determines the flow of resources; whether in the retail line’s “small assistants” or the wholesale line’s “small assistants,” AI no longer waits for business demands but proactively embeds and reshapes business processes.

Local banks are also stepping up efforts. Several banks that have disclosed annual reports have emphasized their AI strategies. Chongqing Bank has created the “Chongyin Xiaoi AI” brand application, becoming one of the first city commercial banks in the country to achieve “privatization + adaptation to financial scenarios” with large models; Qingdao Bank has formulated the “Qingdao Bank Digital Transformation Three-Year Strategic Plan,” which mentions the creation of “two major intelligent engines” for AI capabilities and data value; Ruifeng Rural Commercial Bank has also clarified that by 2025, it will build a bank-wide AI platform based on an open-source framework, forming an intelligent application ecosystem covering major business lines, with AI capability development entering a stage of large-scale application.

AI has also become a high-frequency term in earnings press conferences. Looking to the future, regarding the key tasks for the next phase of building a “digital intelligent ICBC,” ICBC Vice President Zhao Guide pointed out the continuation of the “Leading AI +” action, focusing on four aspects: intelligence, wisdom, intelligent computing, and intelligent sharing; innovatively creating financial intelligent entities, shifting the technology role from back-end support to front-end driving; and accelerating the construction of a “one customer, one advisor” service model, allowing AI to be the most direct bridge connecting banks and customers.

Bank of Communications Vice President Qian Bin emphasized promoting the transition of AI from point applications to comprehensive integration. He proposed strengthening the construction of technological capabilities, deepening service business with employees, upgrading service markets with customers, and improving intelligent risk control levels, clearly indicating that AI has deeply embedded itself in the top-level design of banks, becoming a new productive force driving cost reduction, quality improvement, and efficiency enhancement.

Ping An Bank President Ji Guangheng established digital employees, precise marketing, and precise risk control as the three main focuses, clearly stating the need to strengthen the technological data foundation capabilities, deepen main data management and external data applications, and evolve from human-machine collaboration to intelligent decision-making and automated execution models.

Wu Zewei, a special researcher at Su Commerce Bank, pointed out that AI’s autonomous decision-making, real-time responsiveness, and intelligent learning capabilities will comprehensively reshape banking business models. This includes upgrading customer experience, where AI redefines the connection between banks and customers through multimodal interaction and personalized services, achieving full-cycle customer companionship, personalized wealth management, and real-time anti-fraud monitoring; upgrading risk management, where AI can shift risk control from “post-response” to “real-time interception + predictive warning,” achieving innovations in credit assessment, complex fraud identification, and compliance automation, building a full-process defense network; and upgrading operational efficiency, where AI drives banking business processes to evolve towards “zero contact” and “adaptive,” releasing organizational productivity and achieving process automation, scientific decision-making, and organizational knowledge evolution.

These challenges await resolution

From structural evolution, comprehensive integration, to intelligent decision-making, the AI race in the banking industry has entered deep water.

As banks begin “self-evolving,” what we observe is not only an increase in efficiency but also a return to the essence of financial services. The era of “human sea tactics” in banking is now gone, and the leaders of major banks paint a similar picture. While the implementation of technology and application scenarios continues to accelerate, it is crucial to note that real-world challenges such as data silos and privacy protection still urgently need to be addressed. In addition to these two major issues, banks also face multiple challenges in the deployment of artificial intelligence, including technology adaptation, talent shortages, and regulatory compliance.

How to ensure that technology applications are safer and more controllable has become the primary consideration in the digital transformation of the banking industry. ICBC President Liu Jun candidly stated the preconditions for technology applications at the earnings meeting. He noted, “The technology used by ICBC is relatively new, but this technology must be validated in the market and verified internally with strong capabilities; otherwise, we dare not hastily place this technology on the system, as protecting customer privacy and information security is the bank’s most important responsibility.” Liu Jun emphasized, “Therefore, incorporating advanced technology into operational processes must be preceded by system validation.”

CITIC Bank Vice President Gu Lingyun emphasized, “To make the safety barrier more robust, we must appropriately layout intelligent computing power in advance, introduce new safety technologies, and ensure that AI applications are safe, credible, and controllable.”

Zhao Guide also mentioned the need to enhance governance efficiency and construct a security prevention and control system covering the entire AI application chain, effectively covering areas such as technological infrastructure security, data security, model security, and application security.

According to Dong Ximiao, chief researcher at Zhanglian, the application of artificial intelligence not only drives positive transformations at the business, organizational, and cognitive levels but also brings new issues in technology, regulation, and talent. At the technical level, the “data silos” formed under the fragmented data ecosystem on the data side may lead to model biases, and the protection of data privacy and security during the training process is also an urgent issue to be resolved; the opaque model decision-making process on the algorithm side and the “hallucination” risks of generative artificial intelligence make applications more challenging; threats to cybersecurity have also increased. At the regulatory level, on the one hand, the current financial regulatory system is mainly designed for traditional business models and lacks effective regulatory measures for emerging businesses driven by artificial intelligence technology; on the other hand, multinational financial institutions face compliance challenges arising from differences in regulatory standards across various jurisdictions.

“From China’s actual situation, there are significant differences in scale and business practices between large financial institutions and small to medium-sized financial institutions. Therefore, the paths and strategies for applying artificial intelligence in different financial institutions are likely to be completely different.” Dong Ximiao suggests that large financial institutions should evolve from “tool empowerment” to “value reconstruction,” focusing more on business restructuring, process reengineering, and organizational transformation to create new products, new models, and new business forms. Small and medium-sized institutions, lacking trial-and-error capabilities, should not blindly pursue new hotspots and high technologies but should focus on development direction and business priorities based on their resource endowments, taking a differentiated and specialized approach to transformation and development. At the same time, large financial institutions should take on more responsibility for leading and empowering, becoming “pathfinders” in the research and application of artificial intelligence technology, while also releasing redundant technological capabilities and talent; small and medium-sized institutions should maintain an open and cooperative attitude, collaborating with leading financial institutions or external technology companies, integrating technology ecosystems around business-focused high-frequency scenarios, and accelerating the exploration of “business-technology integration.”

Beijing Business News reporter Song Yitong