Agentic AI

Agentic AI solutions have evolved faster than expected and are already helping banks in the Asia-Pacific region drive efficiency, enhance compliance, and improve customer experience. The first questions banks must consider are where and how to start using Agentic AI. Across the Asia-Pacific region, we are seeing several key use cases emerge:

Frontline services have seen the most visible impact of Agentic AI. Advanced chatbots and virtual assistants are no longer limited to scripted, rudimentary responses. These systems now interpret context, draw from policy databases, and deliver accurate, tailored information to customers.

AI-driven tools are increasingly being used to support customer experience staff. These systems provide real-time prompts, cues, and insights, enabling employees to handle inquiries more effectively and efficiently.

Compliance remains a critical area for AI adoption. Banks are leveraging Agentic AI to embed compliance by design into their processes, automate Know Your Customer (KYC) and anti-money laundering (AML) checks, and monitor risks in real time. These use cases are increasingly helping banks to meet their compliance obligations, a key focus for regulators.

In Hong Kong, while banks remain cautious about potential risks, they are generally enthusiastic about Agentic AI and are adopting it at an increasing pace. There is a growing understanding that Agentic AI should be integrated with broader digital transformation efforts, and banks are using the emergence of this new technology as an opportunity to reassess their technology roadmaps, and to refocus on short term (three-year) horizons.

For many multinational banks, this shift is being guided by their global technology strategies, while local banks are leveraging initiatives like the HKMA’s Generative AI Sandbox to explore use cases. At the same time, the local vendor ecosystem is growing, supported by government subsidies and regulatory initiatives, offering banks a broader range of AI solutions and reducing dependence on global vendors.

Agentic AI will ultimately permeate every department within a bank, from customer-facing operations to back-office functions. For example, dynamic product offerings and real-time pricing adjustments—currently more common in the insurance sector—are just two examples that could become increasingly common in banking in the coming years.

As the number of use cases grow, banks must adopt a strategic approach to ensure they are making the right investments. Here are some key actions banks should consider:

Banks should begin by identifying the specific tasks they want AI to perform, the outcomes they aim to achieve, and the role of humans in the process. This requires upfront design thinking to align AI implementation with broader business goals.

The quality and availability of data are critical to the success of AI initiatives. Banks need to assess their existing datasets, ensure they are clean and structured, and establish robust data governance frameworks.

Due to regulatory considerations, banks are naturally conservative about adopting new and rapidly changing technology. Many banks are currently experimenting with AI in isolated use cases. The challenge lies in scaling these solutions across the enterprise while maintaining safety and compliance. Engaging in regulatory initiatives, such as the Hong Kong Monetary Authority’s (HKMA) Generative AI Sandbox, provides a secure environment to test and refine AI solutions.

AI implementation is resource-intensive, so banks must prioritize use cases that deliver tangible business value. Return on investment (ROI) analyses should guide decisions to ensure investments are directed toward impactful solutions.