Beyond Automation: The Agentic Shift
Since assuming the role of CEO at DBS Group in March 2025, Tan Su Shan has steered Southeast Asia’s largest bank toward a radical technological evolution. Moving past the industry-standard focus on simple chatbots and basic automation, the 58-year-old leader is embedding generative and agentic AI across the bank’s operations. As of late 2025, the organization has deployed over 2,000 AI models and empowered staff to create approximately 26,000 personal AI agents.
Tan describes this transition as “frontier stuff,” aiming for a future of “agentic commerce.” In this model, banking applications may become secondary as AI agents—software capable of reasoning, planning, and executing tasks—interact directly with the bank’s systems to conduct payments, wealth transactions, foreign exchange, and loan processing. This shift represents a move from passive digital banking to proactive, autonomous financial services.
Operational Integration and Economic Impact
The internal platform “DBS-GPT” serves as the backbone of this strategy, allowing employees to conduct research, access policies, and build tailored AI assistants. The adoption has permeated critical departments: corporate bankers utilize AI for drafting credit memos, while wealth management teams leverage the technology to expedite compliance checks and customer onboarding. DBS reported that these data analytics and AI initiatives generated an estimated S$1 billion (US$780 million) in economic value in 2025.
Tan emphasizes that the bank’s current success is grounded in lessons learned from a failed 2014 initiative to build a wealth advisory platform. That experience led the bank to prioritize internal applications that deliver immediate business value before pushing for customer-facing solutions. Currently, DBS is testing payment capabilities for AI agents through partnerships with Visa and Mastercard, positioning itself to remain the “bank of choice” as agentic commerce matures.
The Human-Centric AI Philosophy
Despite the aggressive push into AI, Tan maintains a philosophy of human-employee augmentation. The bank has invested heavily in workforce training, providing access to over 10,000 online courses, hackathons, and technology-focused workshops to help staff adapt to an AI-enabled environment. Tan argues that AI should handle data-heavy tasks, while human relationship managers focus on long-term client connections.
Governance remains a cornerstone of this strategy. Every agent deployed is subject to strict oversight, control, and policy frameworks to manage cybersecurity, data privacy, and trust. Tan has been vocal about the necessity of human oversight for critical financial decisions, such as credit assessments, ensuring that technology serves as a tool rather than a replacement for professional judgment.