DBS Group CEO Tan Su Shan has a message for the banking industry: the more AI you use, the cheaper it gets—provided you embrace what she calls the 'paradox of token spend.' In an exclusive interview with Bloomberg's Haslinda Amin, the head of Southeast Asia's largest bank outlined a future where AI costs decline as adoption scales, even as the broader industry grapples with rising AI expenditures. Her remarks come as DBS reveals that its AI and data analytics initiatives have already generated approximately $1 billion in value, positioning the bank as a regional leader in financial technology innovation.

The 'Token Paradox' Explained

Tan's thesis is counterintuitive at first glance. In the world of generative AI, costs are typically measured in tokens—the fragments of text or code that models process. While token prices have been falling due to increased competition and efficiency gains, many enterprises report that their total AI spending continues to climb. TechNewsWorld recently highlighted this tension in an article titled 'AI Costs Continue to Rise Despite Falling Token Prices,' pointing out that higher usage volumes often outpace per-unit price drops.

Tan, however, sees a different trajectory. She argues that as DBS scales its AI deployments, the cumulative benefits—automated processes, faster decision-making, and personalized customer interactions—will outstrip the incremental costs. 'The paradox of token spend is that the more you use, the more value you extract, and the unit economics improve,' she said, according to Bloomberg. This perspective is gaining traction among financial institutions that are moving beyond pilot projects to enterprise-wide AI adoption.

$1 Billion AI Impact and Agentic AI Outlook

DBS's confidence is backed by tangible results. The bank's AI and data analytics initiatives have generated an estimated $1 billion in value, a figure that emerged from the bank's internal assessments and was reported by Yahoo Finance Singapore. This includes revenue uplift, cost savings, and risk reduction across areas like fraud detection, customer service, and credit scoring.

Looking ahead, Tan predicts the rise of 'agentic AI'—autonomous systems that can execute multi-step tasks without constant human oversight. 'We are moving from assistive AI to agentic AI, where the bank can handle complex workflows in real time,' she explained. DBS has been building proprietary AI capabilities and partnerships to ensure it remains at the forefront of this shift. The bank's open architecture strategy is central to this vision. Tan emphasized that DBS avoids being 'too wedded' to any single model or provider, maintaining a flexible technology stack that allows it to adopt the best AI tools as they emerge.

“The paradox of token spend is that the more you use, the more value you extract, and the unit economics improve.”

Taiwan: The Next Growth Frontier

While AI captures headlines, Tan identified Taiwan as the bank's most exciting market opportunity over the next two to three years. The island is a critical hub in the global semiconductor supply chain, and DBS has been expanding its presence there to serve both corporate clients and wealth management customers. The bank's regional network, combined with Taiwan's tech-driven economy, offers a promising avenue for growth as DBS looks to diversify beyond its home market.

This strategic focus comes as DBS continues to strengthen its position as Southeast Asia's largest bank by assets, with a market capitalization exceeding $100 billion. Under Tan's leadership, the bank has doubled down on digital transformation, a move that has resonated with investors and customers alike. Forbes recently profiled Tan, noting that she is steering DBS into the AI era with a blend of pragmatism and ambition.

Contrasting Views: The AI Cost Debate

The tone from Singapore is optimistic, but the broader industry remains split. TechNewsWorld's report points out that even as token prices decline, the total cost of ownership for AI systems—including infrastructure, talent, and compliance—is escalating. Enterprises are discovering that deploying AI at scale requires significant upfront investment, and the payoff may take years. This is particularly true in regulated industries like banking, where model risk management and explainability demand additional resources.

However, proponents like Tan argue that the long-term value outweighs the short-term costs. DBS's $1 billion figure suggests that the bank is already seeing a strong return on investment. Analysts note that banks with large customer bases and rich data troves are best positioned to monetize AI, while smaller players may struggle to keep pace.

How Different Outlets Frame the Story

  • Bloomberg Markets focused on Tan's 'token paradox' and her open-minded approach to AI vendors, highlighting her prediction of falling costs.
  • Yahoo Finance Singapore emphasized the $1 billion value generated from AI and data analytics, underscoring the concrete financial impact.
  • Forbes framed the story as a leadership milestone, profiling Tan as the CEO leading Southeast Asia's largest bank into a new technological era.
  • TechNewsWorld provided a counterpoint, warning that despite falling token prices, overall AI costs continue to rise for many enterprises.

Implications for Banking and Beyond

DBS's experience offers a blueprint for how financial institutions can navigate the AI revolution. The key takeaways:

  • Embrace open architecture to avoid vendor lock-in and adapt to rapid model evolution.
  • Focus on measurable value—DBS's $1 billion figure demonstrates the importance of tying AI initiatives to bottom-line outcomes.
  • Plan for agentic AI—autonomous systems will define the next wave of productivity gains.
  • Look beyond core markets for growth, as DBS is doing with Taiwan.

As the AI cost debate continues, Tan Su Shan's 'token paradox' may become a defining philosophy for banks seeking to scale AI profitably. Whether the paradox holds across the industry remains to be seen, but DBS's early results are encouraging. For now, the bank's CEO is confident that the future of finance is intelligent, efficient, and—above all—AI-driven.