Not only improving efficiency, but research shows that financial institutions view AI as a strategic transformation engine.

On March 17, PwC released the report “AI Fuels a Fresh Upgrade for Financial Services in Mainland China and Hong Kong” (hereinafter referred to as the “report”), noting that artificial intelligence has moved from experimental pilots to large-scale applications. According to the survey in the report, 76% of financial institutions plan to use AI to drive business strategy transformation and help open up entirely new revenue streams.

The report shows that, through AI investment, the surveyed institutions have achieved an initial return of 11%-15%. Meanwhile, 76% of the surveyed institutions say they are currently willing to accept an investment return of less than 10%, thereby accelerating the realization of AI strategic initiatives. For institutions, while they focus on short-term gains, they value the long-term impact of AI even more—namely, enhancing their market position, expanding strategic development space, and creating new growth opportunities.

Respondents said that the investment returns brought by their AI projects mainly come from multiple areas, including reducing risk losses, improving compliance effectiveness, increasing revenue, and cutting costs. In sharing specific case examples, PwC said that one bank reduced risk losses dramatically by changing a step in its workflow—from previously conducting sampling inspections to now using AI technology for comprehensive inspections.

Five major application scenarios—customer service/chatbot deployments, investment and asset management, fraud detection, predictive analytics and modeling, and back-office process automation—are delivering measurable investment returns and are rapidly becoming key development areas for enterprise-grade AI applications. As one Hong Kong bank executive put it, “It’s not only about pursuing efficiency improvements through AI—we also hope AI will help create new value propositions and business models that emerge in the market.”

At present, different industries have different focuses in their deployment and application of AI. Ni Qing, Partner-in-Charge of Asset and Wealth Management in PwC China’s Mainland business, said, “In banking, the focus is on risk control, anti-money laundering, and compliance responsibilities. In insurance, the emphasis is on enhancing agents’ capabilities, customer service, and claims handling. In the asset and wealth management industry, AI is used in investment and portfolio management, as well as data and market analysis.”

Even so, the level of investment in AI technology remains the core issue. The report shows that for 61% of financial institutions, the proportion of AI investment within their technology budgets is less than 10%, and there is a 30%-to-40% gap between actual AI investment and real needs in the industry. Specifically, 68%, 48%, and 60% of the surveyed institutions in the banking, insurance, and asset management (asset & wealth management) sectors respectively said outright that their AI investment does not exceed 10% of their technology budget.

The large-scale rollout of AI still faces multiple constraints. Respondents said that the top three obstacles to increasing AI investment are data availability, regulatory pressure, and the need to prioritize maintaining existing core systems. In addition, talent shortages and rigid organizational structures are the core obstacles preventing enterprises from scaling up AI deployments, and their impact far exceeds problems at the level of budgets or technology.

Looking ahead, the report believes four major shifts will take place in the financial industry over the next five years: first, hyper-personalized services—shifting from standardized products to an AI-driven dynamic real-time service model; second, highly automated processes and optimized decision-making, with AI taking on more decision-making power and becoming a “super collaborator” for humans; third, proactive intelligent compliance—moving from passive responses to embedded, real-time, and pre-transaction compliance; fourth, real-time predictive risk control.

(Editor: Wenjing)

Keywords: