Huawei unveiled its vision for AI in banking at Huawei Connect 2025 in Shanghai, introducing new Ascend chips and Atlas SuperPoDs designed to power digital employees in financial institutions. The event, themed ‘All Intelligence’ and held at the Shanghai World Expo Exhibition and Convention Centre, showcased how AI in banking is shifting from experimental chatbots to fully operational digital colleagues handling real-world tasks. According to company executives, more than 200 financial institutions in China are already deploying these AI-powered agents across lending, compliance, and fraud monitoring operations.
Jason Cao, CEO of Huawei Digital Finance, explained that the company is moving beyond traditional automation tools. He stated that human-machine collaboration is transitioning from people using tools to people working alongside AI colleagues, enabling one person to complete tasks that previously required an entire team.
Digital Employees Transform AI in Banking Operations
The core proposition centers on pre-trained agents that arrive with industry workflows already embedded, eliminating the need for each bank to build proprietary systems from scratch. These digital employees can integrate with legacy banking systems and execute repeatable, rules-based tasks in areas such as risk management and customer engagement. Rather than remaining in pilot phases, Huawei indicates these agents are already deployed in daily operations across Chinese financial institutions.
In the insurance sector, Pacific Insurance has deployed over 100 categories of digital employees, according to the company. In health underwriting specifically, Huawei reports these agents are operating at accuracy levels exceeding 98 per cent, demonstrating their capability to handle complex, high-stakes decision-making processes.
How Financial Institutions Deploy Digital Workforce Solutions
Huawei’s approach addresses a persistent challenge in banking: administrative burden. Cao pointed to sales teams that spend approximately 70 per cent of their time on document preparation rather than client interaction. If AI colleagues can absorb that administrative load, human staff can redirect their focus toward customer relationships and strategic work.
Additionally, the technology extends to risk control functions. The company described how reinforcement learning combined with expertise from experienced bank risk officers has been distilled into more than 1,000 decision chains. These guide AI colleagues through loan approvals and ongoing monitoring, reportedly lifting SME risk identification by over 50 per cent.
Speed and Scale Requirements for Customer-Facing Banking
However, deployment at scale requires exceptional performance. Research cited at the conference indicated that even a one-second delay in mobile banking services results in nearly a quarter of users abandoning the session. Consequently, AI agents must respond in under half a second to function effectively in customer-facing environments.
Meanwhile, Huawei is building an ecosystem of more than 150 partners contributing industry-specific tools and templates. The newly introduced FinAgent Booster platform is designed to help banks assemble and customize digital employees rapidly, then adapt them across different markets and regulatory environments.
Organizational Culture Shifts Alongside Technology
Cao emphasized that the transformation extends beyond headcount reduction or product development. Leading banks are now contemplating how to create an organizational culture of human-machine synergy, he noted. The vision articulated is “one person, one team, one agent,” with digital colleagues embedded across roles from sales to risk management.
In contrast to traditional efficiency metrics, the focus shifts to creating collaborative models where human judgment and machine processing work in tandem. This represents a fundamental rethinking of workforce structure rather than simple automation of existing processes.
Huawei has not announced a specific timeline for broader international deployment of these digital employee platforms, though the company indicated its partner ecosystem continues to expand. Financial institutions considering adoption will likely evaluate regulatory compliance requirements and integration complexity alongside the operational benefits demonstrated in early deployments.





