DBS chief: AI financing repricing is healthy, not alarming

DBS chief executive Tan Su Shan reckons the rethink now under way in AI-related risk pricing is a healthy one. In a Bloomberg Television interview on 12 October, she argued that investors and lenders pushing back on easy assumptions will force more discipline into how artificial intelligence projects get funded.

Investors, she said, want fatter returns before committing capital, and they are now poring over deal structures instead of accepting them at face value. Her reading is that the market has already delivered its verdict, and the scrutiny is welcome because it forces borrowers to demonstrate that spending on AI infrastructure can service debt while still leaving something on the table for shareholders.

DBS itself has arranged roughly S$25 billion of AI-linked financing this year, according to Tan. The bank usually acts as lead manager on syndicated facilities, where a group of lenders shares the exposure and DBS keeps only a small slice on its own balance sheet. Hardware is where most of the money lands: chips, memory and the data centre builds that house them.

There are clear losers in this repricing. Firms with weaker balance sheets risk being marked down by rating agencies as lenders reassess how much AI exposure they will carry. Meanwhile, elevated interest rates and richer bond yields have made giant build-outs costlier to finance, and PwC puts cumulative Asia-Pacific data centre investment at as much as US$8.2 trillion by 2050.

AI has topped Tan's agenda since she took the top job in March 2025, and Global Finance's 2025 award for the best AI-driven bank globally went to DBS. Early results are turning up in faster account openings and quicker fee income growth, though she gave no figures. The wealth business keeps drawing on Asia's wealth creation, deeper capital markets and the shifting of fortunes between generations, with net new money holding up well even when volatility jolts fund flows.

Why it matters for Singapore: DBS is the republic's biggest lender and a benchmark for how the local financial sector underwrites the AI boom, so tighter terms ripple straight into the cost of capital for the region's data centre and AI build-out. The tone from the top also fits Singapore's pitch as a disciplined, well-governed place to scale AI, where financing prudence and innovation are meant to travel together.