Artificial intelligence investment is sustaining global demand, but central banks should not assume the technology has already raised productivity or changed the economic conditions that guide monetary policy, a senior Monetary Authority of Singapore (MAS) official said.
Edward S. Robinson, MAS deputy managing director for economic policy and chief economist, said the AI boom had created a challenge for policymakers as investment and asset valuations rise ahead of clear evidence of economy-wide productivity gains.
“The AI boom is playing out as a race to balance present demand against future supply effects,” Robinson said in remarks at the 2026 Beijing Financial Street Forum on Oct. 20.
Upbeat expectations have driven a surge in AI investment and expansion, while verified productivity gains have been more limited so far, he said.
The boom has become an important source of economic support for Asia, countering some of the effects of trade tensions and higher energy costs.
AI-driven electronics accounted for more than 70% of Asia’s export growth this year, Robinson said.
Growth in AI-producing industries, however, has so far reflected mainly capital accumulation and higher utilisation of equipment, rather than a broad-based increase in total factor productivity.
The wider economic benefits also remain uncertain, despite AI spreading faster than personal computers and the internet did.
Robinson said sustained productivity gains would depend on wider adoption, investment in skills and infrastructure, and changes to how organisations work.
The Organization for Economic Co-operation and Development estimates that AI could raise total factor productivity growth by up to 0.6 percentage points a year, provided adoption is sufficiently broad and supported by appropriate investment, he said.
Such gains could help offset demographic ageing and other structural pressures in Asia.
For monetary policymakers, the immediate concern is that spending on data centres and energy networks is adding to demand before the technology’s productivity benefits are fully established.
Higher asset valuations could provide another boost through wealth effects, Robinson said.
He warned that expectations of large future productivity gains could become embedded in asset prices.
Excessive optimism could encourage leverage and push valuations beyond underlying fundamentals, leaving markets vulnerable to reversals that would affect asset prices, credit conditions, and investment.
Monetary policy should remain appropriately restrictive in the initial stages of the boom, when demand effects dominate, before becoming more accommodative as productivity gains demonstrably take hold, Robinson said.
Policymakers should adopt a gradual, data-dependent approach rather than assume that AI has already changed potential output, the natural rate of unemployment or the neutral interest rate.
The eventual effect on interest rates is uncertain, he added. Sustained productivity growth could raise potential output and the natural interest rate, while uneven gains, job displacement and income uncertainty could prompt households to save more, putting downward pressure on the natural rate.
Robinson called for a broader policy response, including macroprudential measures to guard against excessive borrowing based on uncertain future revenues.
Governments should also support AI adoption through digital infrastructure, skills training and applications that help workers perform more productive tasks, while establishing regulatory safeguards to maintain trust in the technology.
He said AI could ultimately expand human creativity and enable more people to innovate and exercise judgment in their work. But past episodes in which economic expectations outpaced productivity and asset valuations ran ahead of fundamentals offered a reason not to assume the current boom would be different.
The challenge for policymakers is to manage the transition as the demand and supply effects of AI converge, while maintaining price and financial stability, Robinson said.
