Energy Is the Chokepoint for AI's Next Phase, DBS CIO Says
Source: The Business Times
DBS Bank's CIO warns that energy infrastructure has become the primary constraint on AI growth and recommends investors add power-sector exposure alongside technology stocks, highlighting opportunities in Singapore's utilities and clean energy companies.
Energy has become the critical bottleneck holding back the next wave of artificial intelligence expansion, and investors should pivot toward power infrastructure as a core AI-adjacent theme, according to DBS Bank's Chief Investment Office. In its CIO Insights for the second half of 2026, DBS flagged that surging electricity demand from AI data centres is straining grids globally, making energy availability — not chip supply or model capability — the primary constraint on AI's growth trajectory.
The DBS CIO pointed to a structural mismatch: AI data centre power consumption is forecast to grow at a compound annual rate of 25 to 30 per cent through 2030, yet new power generation and transmission infrastructure takes years to bring online. This gap, the bank argues, creates a compelling investment opportunity in companies and assets that sit at the intersection of AI demand and energy supply — from utility providers and renewable energy developers to power infrastructure Reits and grid-equipment manufacturers.
Within Singapore specifically, DBS highlighted opportunities in utilities and companies exposed to clean energy, renewables, and power infrastructure. Singapore's push to expand its data centre capacity — while managing its carbon commitments under the Green Plan 2030 — makes energy infrastructure a strategic national priority as much as an investment theme. The republic's goal of reaching 2 gigawatts of data centre capacity by 2030, paired with its import of low-carbon electricity from regional neighbours, positions Singapore as a test case for balancing AI growth with sustainability targets.
The call comes as AI model training and inference costs continue to fall, widening the range of economically viable use cases. But without commensurate investment in power generation and grid capacity, those efficiency gains risk running into a physical ceiling. DBS's analysis suggests that every new large language model deployment consumes electricity equivalent to thousands of households annually, and that the gap between AI-driven demand and available clean energy supply is widening, not narrowing.
Why it matters for Singapore: DBS's framing reframes the AI investment conversation for a city-state that is both a financial hub and a data centre gateway for Southeast Asia. Singapore-listed utilities, Reits with data centre exposure, and companies in the power infrastructure supply chain stand to benefit as institutional capital rotates toward AI-adjacent energy plays. More broadly, the analysis underscores that Singapore's AI ambitions cannot be decoupled from its energy strategy — and that the next phase of AI competition will be shaped as much by kilowatt-hours as by algorithms.