How DBS CEO Tan Su Shan is reshaping Southeast Asia’s largest bank for the AI age

By Phong Ngo   October 4, 2026 | 04:40 pm PT Google Get VnExpress first in Google Search info See more of the news you trust. Make VnExpress a preferred source to prioritise our updates in your Google search results
DBS CEO Tan Su Shan is bringing AI deeper into her bank, while keeping human judgment, trust and relationships at the heart of its future.

DBS, founded in 1968 to support Singapore's industrialization, has grown alongside the city-state into Southeast Asia's largest bank by assets. More than five decades later, it is betting on artificial intelligence to drive its next transformation.

Tan, ranked the world's sixth most powerful businesswoman by Fortune in 2026, believes the transformation is only beginning.

"Soon the world will morph from digital banking to agentic banking," Tan told The Business Times in an Oct. 1 interview.

From digital banking to agentic banking

Unlike conventional AI assistants that mainly answer questions, AI agents can plan and carry out tasks using digital tools with limited human intervention. In banking, they could eventually handle activities ranging from payments and loans to investments.

DBS is already preparing for that world. Within months of making the capability available in late 2025, employees created 10,000 personal AI agents. The bank divides its approach into personal agents, team agents and more sophisticated enterprise agents designed to handle complex workflows.

Tan uses one herself to track banking news and what other bank CEOs around the world are saying.

At a DBS leadership gathering this year, attended by 200 to 300 senior executives, participants wore T-shirts reading: "Talk is cheap. Show me your agent." They then learned to build one themselves. Today, seven in 10 DBS employees use generative AI tools, including the bank's internal DBSGPT platform and Microsoft Copilot. Together, they generate around 1.8 million prompts a month for research, writing, translation, brainstorming and summarization.

The technology is also moving into more complex banking work. In credit assessment, as many as 70 specialized agents can perform different parts of the process and produce a draft credit memo, which a human credit manager reviews before making the final decision.

DBS says the technology is already producing measurable returns.

AI and data analytics generated about S$1 billion (US$782 million) in economic value for the bank in 2025. In wealth management, AI helped cut onboarding turnaround times by 50% in the first five months of 2026.

"We're beginning to see the green shoots of AI working in generating ideation that generates fees," Tan told Channel News Asia in August. She said AI was helping younger relationship managers develop ideas for clients and allowing the bank to generate more fee income with the same workforce.

Tan Su Shan, CEO of Singapores DBS Group. Photo courtesy of DBS

Tan Su Shan, CEO of Singapore's DBS Group. Photo courtesy of DBS

The transformation follows more than a decade of experimentation. DBS began working with IBM's Watson in 2014 on an AI-powered wealth advisory system, but the technology achieved only about 80% accuracy and ultimately did not work as hoped.

Tan, who joined DBS in 2010 after working at Morgan Stanley and Citi, took over as CEO in March 2025, becoming the first woman to lead the bank. Her rise to the top also put her directly in the path of the disruption she is now asking employees to embrace. On the day she was appointed CEO, someone in the DBS board's WhatsApp group sent her a message: "You know, even a CEO's job can be replaced by AI."

For Tan, the challenge is not simply preserving existing jobs. DBS's approach is to "protect workers, not jobs," retraining employees as AI takes over more routine work and changes the skills the bank needs.

All 40,000 DBS employees are receiving foundational AI training, with 11,000 undergoing deeper training for changing or emerging roles. Tan said AI will also sharply reduce the roughly 3,000 production engineers the bank once needed, requiring employees to work at a higher level.

An AI-enabled bank with a heart

As machines take on more work, DBS faces a more fundamental challenge: how to give AI greater autonomy without compromising the trust and accountability on which banking depends.

DBS sees trust, data and culture as three strategic moats in navigating that transformation. Its customer and transaction data can make AI more useful, while its culture helps employees adapt to the technology and trust becomes increasingly important as agents gain greater autonomy.

To manage the risks, DBS has introduced an agent registry and controls governing what data agents can access, how their actions are tracked and who remains accountable for them. The bank eventually plans to expand beyond internal agents to those that could help customers directly, making safeguards around data, control and accountability even more important.

DBS sees that transformation extending beyond the bank itself as it seeks a larger role in Asia's growth.

"In the next decade, Asia will remain a major player in the global economy. We want DBS to be a trusted partner in that growth, not just keeping pace with technology but using it meaningfully to shape the future of banking," Tan said.

For Tan, however, the answer ultimately lies in the relationships, judgment and trust that technology cannot replicate. "AI doesn't have a heart. Humans do. And we want to put the human in the center of everything we do," she said.

That philosophy underpins DBS's ambition to become what Tan calls an "AI-enabled bank with a heart": using machines to make employees more capable and services more efficient without allowing technology to displace the human qualities on which banking relationships depend.

Her rule for navigating that transformation is simple. "When things are bad, don't lose your heart. When things are good, don't lose your head."

Google Get VnExpress first in Google Search info See more of the news you trust. Make VnExpress a preferred source to prioritise our updates in your Google search results
 
 
go to top