Hong Kong banks expand AI across lending and fraud controls

Virtual banks are using richer data and modern technology stacks to compete in lending and risk management.

Hong Kong banks are expanding AI across lending, customer service, and fraud prevention as adoption broadens beyond early-stage testing.

More than three-quarters of banks are already deploying or piloting AI. Stanley Sum, Partner and Head of Technology of the Greater Bay Area at KPMG China, said virtual banks can differentiate themselves through newer technology stacks that support faster product development and more flexible digital services.

“The advantage isn't simply how much you invest in AI, but how thoughtfully you weave it into your overall customer experience,” Sum said.

In SME lending, AI can help banks assess businesses with limited financial statements or collateral by combining traditional measures with cash flow, trade, e-commerce, and payment data.

Sum pointed to Hong Kong’s Commercial Data Interchange, which had supported more than 95,000 loan applications by March.

He said banks should use richer data to understand business cash flows, monitor borrowers continuously, and expand lending gradually after building experience with simpler working-capital products.

Wider use of AI also raises governance concerns. Sum said banks need to ensure customers can understand how decisions are reached, whilst also testing models for fairness and making sure generative AI produces accurate and suitable advice.

Fraud is another concern as AI-enabled scams become more sophisticated, particularly across digital channels.

“Customers should always know that there are people accountable behind every AI-driven decision,” Sum said.

For virtual banks, the competitive advantage will depend on how well they combine technology, data, and risk controls across core services.

Sum will speak at the Asian Banking & Finance and Insurance Asia Summit in Hong Kong on 5 November.