DBS has added generative AI and agentic AI functions to its virtual assistants DBS Joy and DBS digibot, now reaching more than 10 million customers across Singapore, Hong Kong and Taiwan. The upgrade lets authenticated users direct the assistants to complete simple banking tasks - not just answer questions - reducing routine support queries and freeing human agents for more complex issues.
Business assistant moves beyond Q&A
DBS Joy, the bank's assistant for corporate and SME customers inside the DBS IDEAL platform, can now check whether a payment was completed, confirm if money was transferred to a specific company, or show how much was paid in fees. The assistant acts only when a logged-in user gives the instruction, pulling account and transaction data in real time. The service is live for 350,000 corporate users in Singapore, with a further 100,000 users in Hong Kong due to receive it later.
Demand has climbed sharply. In the first half of the year, chats on DBS Joy in Singapore rose 22% and active users jumped 61%. Customer service calls and emails fell 7% over the same period, while satisfaction scores for the assistant increased 17%. The assistant's ability to handle tasks has helped cut routine calls and emails, a trend that mirrors the growing use of AI for Customer Support across financial services.
Retail and wealth upgrades
DBS digibot serves more than five million retail customers in Singapore and another four million across Hong Kong and Taiwan. In the first half of the year, it resolved nine out of 10 customer queries digitally, without a follow-up call. New generative AI features have been added across all three markets to help the assistant understand requests better and guide customers through routine tasks in more natural conversations.
The assistant is being integrated into digiWealth, the bank's mass-market wealth management platform, where it will answer wealth-related questions and connect customers to their wealth planning manager. Later, DBS digibot will add agentic and voice AI functions for selected journeys - checking card usage, calculating reward points, requesting fee waivers, and blocking or replacing cards. These capabilities show how AI Agents & Automation are reshaping customer service beyond simple FAQs.
Scaling support with AI
The expansion reflects a wider push by banks to use AI in front-line service while routing complex cases to human staff. DBS is adding live chat links to human agents for corporate customers when needed and expanding the assistant's knowledge of SME banking products. Customer support remains available for users who prefer to speak with a service officer.
Derrick Goh, Chief Operating Officer at DBS, said: "The true value of AI lies in delivering meaningful outcomes for customers at scale. Today, DBS Joy and DBS digibot serve more than 10 million customers across the region, enabling us to bring Gen AI and agentic AI into everyday banking interactions. Building on more than a decade of AI innovation, we are taking the next leap forward - moving from helping customers find information to helping them get things done. This is especially significant in corporate banking, where digital assistants remain uncommon and agentic capabilities even more so. By combining our deep digital capabilities, trusted customer relationships and proprietary AI platforms, we are reimagining how banking services are delivered."
DBS expects the virtual assistants to handle more than one million chats a month as the new functions roll out across its customer base.
Why this matters for customer support professionals
DBS's move signals that agentic AI in customer service is moving from experimental to operational at scale. For support teams, these tools can shrink ticket volumes by resolving routine requests autonomously - deflecting password resets, status checks, and simple transactions without agent involvement. The 7% drop in calls and emails at DBS, alongside a 17% satisfaction gain, suggests customers accept and even prefer this model when it works fast and accurately.
The critical design question becomes escalation: when should the bot hand off to a human? DBS's approach - keeping live chat links accessible and training the assistant on product knowledge - points to a hybrid model where AI handles the known, and people manage the unknown. Support professionals will spend less time on repetitive work and more on high-judgment conversations, empathy-driven interactions, and complex problem-solving. The skill set shifts from keyboard shortcuts to coaching and exception handling.
Your membership also unlocks:
