The demand for professionals who grasp both business fundamentals and AI applications is evolving from a desirable bonus into a standard requirement listed in job descriptions. As September commenced, recruitment announcements for campus hires were successively released by ICBC, Agricultural Bank of China, Bank of China, China Construction Bank, and Postal Savings Bank of China; earlier, on August 10th, Ping An Bank had already initiated its online application process, followed by CM BANK on August 28th. With these moves, the autumn recruitment season for national commercial banks has officially commenced.

Data from the third-party platform Bank Recruitment Network indicates that the combined 2027 campus recruitment plans of the four major banks—ICBC, ABC, BOC, and CCB—total approximately 64,000 positions, reflecting a decrease of about 7,000 compared to the previous year. Concurrently, the interim reports of the six major state-owned banks for the first half of 2026 reveal that their active workforce shrank by approximately 30,500 employees since the beginning of the year. Among these, Agricultural Bank of China saw a net reduction of 10,900 staff, while ICBC reduced its headcount by 7,210. While banks continue to hire, they are simultaneously downsizing their existing workforce. For banking professionals, the newly added campus positions signal where each institution plans to invest future resources; conversely, the compressed personnel numbers highlight the risk exposure in existing roles. The total reduction in this year's autumn recruitment among the four major banks amounts to roughly 7,000 positions.

Looking at the overall numbers, the 64,000 planned hires from the four major banks still represent a substantial intake, primarily allocated to branch offices and county-level networks. For instance, ICBC's Guangdong Provincial Branch alone is offering 1,300 positions, with branches in Heilongjiang and Inner Mongolia planning to recruit 697 and 520 individuals respectively. Recruitment has not ceased, yet the overall scale is clearly trending downward. According to statistics from the Bank Recruitment Network, Bank of China's autumn recruitment plan for 2027 stands at approximately 11,200 positions, a cut of around 2,000 from the prior year. This reduction is not confined to recruitment figures. By the end of 2025, the total workforce at the six major banks was 1.8544 million, slightly up by 1,538 from a year earlier; nevertheless, by the end of June 2026, active employees shifted to a net decrease of roughly 30,500, reversing from growth to contraction within just six months. In contrast, joint-stock banks experienced milder adjustments, with Ping An Bank reducing its workforce by 655 employees over the half year, while CM BANK bucked the trend by adding 374 net new staff.

Regarding branch networks, the peak of contraction occurred in 2025, with the industry witnessing a net closure of approximately 2,900 outlets, predominantly among rural commercial banks, rural credit cooperatives, and village banks. Entering 2026, the total number of commercial bank branches has stabilized, with the big state-owned banks reducing their physical presence by only about 49 in the first half of the year. An analysis of annual reports and campus recruitment announcements reveals that workforce reductions are concentrated in roles that are standardizable and susceptible to automation. A frequent pattern of internal mobility over the past two years involves tellers transitioning to positions in lobby management, wealth management, or relationship management.

This year's recruitment wave shows a significant technological tilt, with 506 positions related to technology and an explicit call for individuals who are both AI-savvy and business-minded. In its 2027 campus recruitment, ICBC has listed 'Technology Elite' as one of its six main role categories, with 506 relevant postings out of 2,567 listed on its recruitment page. CM BANK's head office functional positions include a dedicated track for intelligent science research and development, with its announcement explicitly stating an intention to cultivate hybrid professionals who are 'proficient in AI and understand business operations'. SPD Bank has established specialized head office technology reserve positions, with no strict restrictions on academic majors, though applicants with backgrounds in artificial intelligence and data science are given priority. Similarly, Ping An Bank's head office 'Digital Intelligence Finance Trainee' program imposes no major restrictions, favoring candidates with interdisciplinary knowledge, and the role's responsibilities involve integrating data analytics and intelligent technology with traditional financial services.

Technical acumen is also increasingly required for non-technical roles. According to the 2027 campus recruitment notice from CMB Wealth Management, its research and trading positions demand proficiency in utilizing various AI large model tools, familiarity with at least two to three mainstream products, and the ability to deeply embed AI into daily investment research workflows, with a preference for candidates holding interdisciplinary backgrounds in science and engineering. Similarly, Bank of Ningbo's head office recruitment lists two categories: 100 positions for directed financial technology students and 50 for head office trainees, with the former encompassing product manager and functional management tracks. The functional management direction requires an understanding of AI technology principles, including large models, and awareness of AI-assisted programming. The boundaries between technical and non-technical roles are blurring—product and functional directions now appear within technology tracks, while AI tool requirements are embedded in non-technical job descriptions.

Amidst workforce reductions in the first half of 2026, compensation has paradoxically increased. The combined salary expenses of 42 listed banks reached 484.755 billion yuan, a 3% year-on-year growth. According to annual reports and interim statements, the six major banks saw their technology talent pool expand by approximately 35,000 in 2025, reaching 135,900 by year-end. CM BANK employs 11,000 R&D personnel, accounting for 8.98% of its total staff, while SPD Bank's parent company has 6,340 technical staff, exceeding 10% of its workforce. Even as overall headcount contracts, technology departments continue their expansion. The premium for high-end AI talent is evident: in March of this year, Bank of Chengdu advertised an annual salary ranging from 500,000 to 800,000 yuan to recruit an artificial intelligence expert for its head office information technology department, tasked with leading the bank's overarching AI architecture design. For existing employees in business roles, the expectation to 'understand business and utilize AI' is shifting from a competitive advantage to a baseline qualification documented in job descriptions.

AI is now also conducting the initial screening of candidates, with the first test administered by the model itself. During this autumn's recruitment, CM BANK explicitly designated AI interviews as a mandatory step in its announcement, while Ping An Bank began sending out initial AI interview notifications in late August. Although state-owned banks, joint-stock banks, and city commercial banks had previously integrated AI interviews into their preliminary screening processes, this year marks the formal codification of this practice into published rules. AI interviews transitioning from an unobtrusive procedure to a stated term suggests that AI's transformative impact extends beyond business operations and into the very fabric of human resource management. The initial filtering is delegated to algorithms, with human reviewers only evaluating those candidates who pass the AI's selection. This recruitment philosophy reflects the large-scale adoption of AI applications.

By the end of June, ICBC's 'ICBC Intelligence' large model technology framework had been deployed across more than 600 application scenarios. China Construction Bank reported generative AI coverage across more than 600 application scenarios, while China CITIC Bank accumulated over 1,900 AI service scenarios. The impact on workload is direct and measurable. ICBC's 2025 annual report disclosed that its AI digital employees undertook work equivalent to 55,000 full-time employees working for an entire year. In the first half of 2026, ICBC's AI systems automated 100 million manual entries in its operations management domain, processing vouchers 25 times faster than humans. Its remote authorization intelligent review handled approximately 50% of the bank's total transaction volume and completed full quality inspections across all business without new hires, achieving an accuracy rate exceeding 98%. Meanwhile, CM BANK reported that AI applications generated the equivalent of 13.88 million hours of manual labor during the same period. The campus recruitment roster serves as a roadmap of resource allocation, revealing where banks are headed: total numbers are tightening, structure is shifting, and the workload absorbed by AI is already quantified in hundreds of millions of entries and half of all banking transactions. The direction of candidate screening and workforce deployment is aligned, and where jobs move ultimately depends on how work is reorganized and reassigned.