Artificial Intelligence in the Financial Sector: Background, Current Practices, Challenges, and Future Trends

Introduction – Recent advances in data, models, and computing have accelerated AI development, making it a key driver of technological and industrial revolutions. In finance, AI is becoming essential as banks digitize and complexify their services.

Expert Profile – Li Jinlong, senior manager of the AI Lab at China Merchants Bank, has led projects in capital agreements, data standards, big data, blockchain, and AI, earning multiple national awards and contributing to dozens of academic papers and patents.

AI in Finance: Background and Significance – Banking’s high digitalization and complex operations make it a prime candidate for AI empowerment. Networked and digitalized processes create data loops that enable AI models to self‑learn, improving system efficiency and customer experience.

Current State of AI Applications in Finance – AI drives intelligent finance by automating tasks, enhancing productivity, and creating new business models. Large language models (LLMs) offer human‑like interaction, document generation, reasoning, and coding capabilities, opening new competitive fronts. Banks are applying AI in wealth management (report generation, alpha factor extraction), customer service (AI assistants, intelligent outbound calls), marketing (script generation), operations (automated quality checks, multimodal RPA), system development (code generation, quality monitoring), and office work (meeting notes, report drafting).

Case Study: China Merchants Bank – Since establishing its AI Lab in 2017, the bank has built over a hundred AI‑enabled scenarios across customer service, operations, and investment research, deploying products such as AI XiaoZhao, text‑based chatbots, intelligent outbound calls, and collaborative AI assistants.

Challenges Facing AI in Finance – 1) Incomplete data supply ecosystems: limited data sources, heterogeneous unstructured data, and privacy concerns hinder data quality and usage. 2) Model risk and compliance: high accuracy demands, black‑box opacity, and fairness issues complicate regulatory oversight. 3) Massive compute requirements: training and serving large models demand resources beyond most banks’ budgets.

Future Trends – AI will further personalize customer services (tailored investment advice), automate operational management (human‑machine collaboration, intelligent task replacement), and enhance risk management (knowledge graphs, big‑data analytics for fraud detection and risk prediction). While opportunities abound, attention to privacy, algorithmic bias, and sustainable governance remains essential.

Conclusion – AI’s deepening penetration in finance promises innovation and transformation across the industry, but success depends on addressing data, model, and infrastructure challenges while ensuring responsible deployment.