Huawei Charts Path to Agentic Banking With Open-Source AI Models
Huawei announced six initiatives at its Shanghai finance summit to help banks deploy AI at scale, moving beyond digital systems toward autonomous banking agents that handle customer service and risk decisions.
The company unveiled upgraded versions of its Financial Data Intelligence and Digital CORE solutions, along with new AI infrastructure designed for financial institutions. The announcements came at the Huawei Intelligent Finance Summit 2026, which drew over 800 representatives from financial firms across 60+ countries.
The shift to agent-based banking
Banks are moving from rule-based systems to AI agents capable of independent decision-making. This means customers interact through language-based interfaces rather than traditional menus, while employees use AI assistants to handle routine work.
Huawei's strategy centers on open-source foundation models paired with private deployment. This approach addresses a core problem: banks need AI that works within their regulatory constraints and data security requirements, not generic cloud solutions.
Jason Cao, CEO of Huawei Digital Finance, said the company has spent 16 years building financial-grade infrastructure and is now applying that expertise to AI deployment. "Building core advantages based on open-source models is becoming the strategic direction for global financial institutions," Cao said.
Four focus areas for AI deployment
Huawei and its partners developed nine AI agent solutions targeting four business domains:
- Intelligent interaction-customer-facing AI that personalizes responses based on user history
- Efficient operations-automating routine tasks across departments
- Intelligent risk control-detecting fraud and predicting defaults faster
- Revenue growth-identifying upsell opportunities and improving customer retention
In practice, this means banks can detect fraud in 30 milliseconds instead of hours, and analyze risk cases 40 times faster using AI-driven analysis rather than manual review.
Data architecture gets an overhaul
Financial Data Intelligence 6.0 reorganizes how banks store and use data. The system processes both structured data (transactions, customer records) and unstructured data (documents, videos, emails) in a single platform designed for AI consumption.
Banks using this approach have improved risk identification accuracy by 25% and boosted customer retention through real-time processing, according to Huawei.
The solution includes automated data governance in partnership with firms like Keyrus and Sunline, reducing manual data management work.
Core banking modernization accelerates
Digital CORE Solution 6.0 helps banks migrate legacy mainframe systems to cloud-native architecture. Huawei's AI-powered code translation tool achieves over 90% adoption rates, cutting planning and design cycles by more than half.
The company has already modernized core systems at over 150 financial institutions globally. New partnerships focus on credit card processing, central bank payments, and insurance platforms.
TaiShan general-purpose processors now run containerized applications at 80% of Chinese banks and institutions across eight countries, offering an alternative to traditional x86 servers.
Building resilient AI infrastructure
Huawei introduced the Atlas 850E SuperPoD, designed as enterprise-grade AI computing hardware for financial institutions. The system pairs high-performance processors with networking optimized for AI workloads.
The company is also building hybrid architectures that split AI processing between cloud and on-premises systems. This approach keeps sensitive data local while using cloud resources for non-critical tasks, reducing token costs-a key expense metric for large language model operations.
Talent development becomes a bottleneck
Huawei committed to training over 10,000 professionals in finance and AI over the next three years. The shortage of people who understand both banking operations and AI implementation is slowing adoption across the industry.
The company has served 7,100 financial institutions across 80+ countries over its 16-year involvement in finance. Its strategy now focuses on making AI deployment faster and cheaper through engineering practices that work at scale.
For finance professionals, the practical implication is clear: AI agents will handle routine decisions, but human judgment remains essential for strategy, exceptions, and trust-based decisions. Banks that train staff to work alongside AI agents-rather than viewing them as replacements-will capture more value from the technology.
Learn more about AI for Finance and Generative AI and LLM to understand how these technologies are reshaping financial services.
Your membership also unlocks:
