11:42 pm - February 4, 2026
Banks worldwide are struggling to move artificial intelligence from the pilot phase into full production, even as competitive pressure mounts. Research indicates that only 11 per cent of banks have deployed generative AI in production environments, despite 43 per cent actively working toward rollout, according to a Temenos survey. More than 80 per cent of banking executives now agree that institutions failing to implement AI will fall behind their peers.
Huawei Digital Finance recently launched FinAgent Booster (FAB) at the Huawei Connect 2025 conference in Shanghai to address this deployment gap. CEO Jason Cao said the platform aims to bridge the divide between promising AI demonstrations and operational banking systems by providing ready-made workflows and integration tools drawn from real-world financial use cases.
Why AI in Banking Remains Stuck in Development
The disconnect between ambition and execution stems from missing infrastructure rather than lack of interest. Cao explained that while AI models exist and banks want the technology, most institutions lack the engineering foundations, workflows and supporting systems needed to scale solutions beyond testing environments.
“AI in finance is already evolving from an assistant role to core business scenarios such as customer engagement, risk management, even end-to-end processes,” Cao said. However, he noted that from an institutional perspective, critical components remain absent, making deployment far more complex than initial demonstrations suggest.
How FinAgent Booster Tackles Deployment Challenges
FAB delivers more than 50 scenario workflows and demonstrations based on actual banking operations, giving institutions a tested foundation instead of requiring custom development from scratch. The platform includes over 150 micro-component plug-ins designed to connect AI agents with both modern systems and legacy infrastructure that still powers much of banking operations.
This approach targets mid-sized banks particularly hard-hit by the modernization challenge. While large institutions possess dedicated AI teams and substantial budgets for lengthy deployments, smaller players face mounting pressure to modernize with significantly fewer resources and technical staff.
Balancing Legacy Systems With New Technology
Every financial institution operates with unique combinations of workflows, compliance requirements and existing platforms, making standardized solutions difficult to implement. Cao acknowledged this complexity but said FAB specifically addresses integration challenges through its plug-in architecture that works across both AI-native applications and older systems.
“For the AI-native ones it’s faster, but you still have to connect with legacy,” Cao explained. The micro-component plug-ins enable AI agents to interface with traditional banking platforms based on Huawei’s accumulated engineering experience across multiple deployments.
Additionally, the platform incorporates lessons learned across different markets and regulatory environments. Huawei reports that solutions tested in one region can be refined and redeployed elsewhere, reducing redundant development work when banks expand AI implementations globally.
Speed Over Perfection in AI Adoption
Cao emphasized that rapid iteration matters more than waiting for perfect plans when deploying AI in banking. Since no proven roadmap exists for enterprise AI adoption, institutions benefit from early experimentation even when initial efforts produce mistakes or require adjustment.
“There’s not a clearly defined path, so people are exploring,” Cao said. Meanwhile, FAB’s technical performance includes over 90 per cent accuracy in intent recognition for customer-facing agents while maintaining millisecond response times, according to Huawei.
Managing Expectations for Returns
Banking leaders frequently demand immediate return on investment from AI projects, but Cao cautioned against short-term thinking that could undermine long-term strategic value. He compared AI adoption to raising a child, noting that demanding immediate returns from developing technology risks creating harmful organizational expectations.
“We definitely should not underestimate the value AI can bring in the long run, but we also cannot overestimate what it can do in the short run,” the Huawei executive said. In contrast to quarterly performance metrics, AI deployment requires sustained investment before delivering transformative business outcomes.
The banking industry faces continued uncertainty around AI deployment timelines and standardized approaches. However, institutions that begin implementation now while building on shared technical foundations are positioned to gain competitive advantages as the technology matures and proven use cases emerge across the financial sector.
