AI making banking more accessible, professional

Artificial intelligence's potential to make financial services more professional and accessible drew wide attention at the International Forum for China Financial Inclusion 2025, which closed on Thursday in Beijing.

Ma Xiaohang, general manager of MYbank's information technology department, told the forum that advances in large AI models — particularly in reasoning, tool use and data processing — are enabling breakthroughs. "This gives us confidence that more professional financial services can indeed become more inclusive," he said.

Earlier this year, MYbank, a digital bank using data-driven lending to serve small and micro enterprises, unveiled its AI banking strategy, aiming to act as "the CFO for every small business" while equipping each bank employee with an intelligent assistant.

MYbank's AI applications over the past year have focused on three areas: risk management in lending, customer operations and services, and wealth management. The bank is developing AI assistants for credit approval, capable of streamlining industry research, due diligence and loan assessment.

"In the past, developing industry knowledge and building a risk framework could take experts one to two months. With AI, we can now automate data collection, build analytical frameworks and identify the key risk factors of an industry in minutes," Ma said. He added that AI-powered due diligence tools, including video and photo verification, are helping the bank better assess client operations and risks in the absence of physical branches.

MYbank primarily serves small and micro enterprises with loans under 2 million yuan ($280,370). Ma noted that the bank's AI-generated credit limits now align with expert assessments in roughly 90 percent of cases. He stressed that computing power, proprietary data, and specialized models will be essential for applying large AI models in financial services.

Bank of Jiangsu is also pushing forward with AI adoption. Since early 2023, the bank has explored building large models, and it accelerated deployment after the release of an open-source reasoning model developed by Chinese AI startup DeepSeek in January.

Shi Cheng, general manager of Bank of Jiangsu's inclusive finance and technology finance departments, said AI is mainly being applied in business strategy, risk control and operations. For example, AI now identifies low-risk, high-value small business clients by analyzing years of customer data.

"We have AI analyze the characteristics and profiles of low-risk, high-value small and micro enterprise clients identified through each data run. Then, we broaden our perspective to the entire market, exploring new avenues while building upon our existing business direction," he said.

In risk prevention, the bank uses AI to detect potential fraud scenarios by learning from past cases. In operations, an AI-based contract review reduces processing time from minutes to seconds.

Liu Xiaochun, affiliate professor at the Shanghai Advanced Institute of Finance at Shanghai Jiao Tong University and vice-president of the Shanghai Finance Institute, said that despite the recent surge of interest in AI, financial institutions have remained calm, unlike the more impulsive embrace of the earlier Internet Plus wave. He noted that institutions are carefully examining the specific features of different AI technologies and assessing whether they can address issues of efficiency or cost.

Liu stressed the importance of understanding the relationship between AI and human beings, emphasizing that people must always come first, since AI is ultimately developed by humans.

"When a bank adopts AI, it is the bank that decides how to use it and what problems AI should solve," he said. "Technology can reduce the need for manual work, but that does not mean people are unimportant."

Reflecting on financial innovation, Liu said that despite evolving from internet finance to fintech, digital finance, and now AI, the problems they address are quite similar and the challenges remain largely the same — especially the "last mile". He cautioned against exaggerating AI's capabilities, stressing that it is people who ultimately solve problems, and only with clear understanding can AI be used effectively.

jiangxueqing@chinadaily.com.cn