Lawrence Wong urges Singapore to use AI boom to build jobs before an uncertain correction

Singapore Prime Minister Lawrence Wong urged the country to use the AI boom to build capabilities and better jobs, warning at the Forbes Global CEO Conference in Singapore on Thursday, 8 October, that a correction will eventually come. He said the current demand has benefited Singapore’s chip industry, but the timing and effects of any reversal are unknown.

Speaking with Forbes Media chairman and editor-in-chief Steve Forbes at Shangri-La Singapore, Wong said: ‘No boom is forever. At some point, there will be a correction.’ He added that he did not know when or how it would happen, or who its casualties would be. His proposed response was to make use of the present demand while it lasts.

How Singapore is benefiting from AI demand

Wong said Singapore does not make the most advanced AI chips. It does, however, host major manufacturers including GlobalFoundries and Micron, which he said make memory and specialty chips now in demand because of AI. He also pointed to a wider semiconductor ecosystem of precision engineering businesses and equipment suppliers.

The benefits extend beyond chipmaking, according to Wong. He cited positive spillovers into logistics and professional services. His aim, he said, is to turn the current window of external demand into new capabilities and better jobs for Singaporeans. He did not put a date on a correction or identify which parts of the economy might be most exposed.

Why conference speakers warned about AI spending

Kande said many executives treated AI as a technology project rather than a change to the business itself. In his account, that left many uses focused on productivity gains absorbed within organisations. He also said some companies lacked the digital systems and governance needed to use their data effectively. Those observations concern the returns companies are seeing now, not a forecast of when an investment correction might arrive.

Banyan Group chairman Ho Kwon Ping raised a different concern at the same session: the physical assets being built for AI. The Business Times reported his warning that data centres across Asia are often financed by large funds and may depend on contracts he considered relatively easy to break. He questioned which countries could be left with stranded assets if the data-centre boom fades. That was a warning about a possible outcome, not a report that such a bust has occurred.

Ho also said AI’s benefits were clear to him, while distinguishing those benefits from the risk of a physical and investment bust. Taken together, the conference remarks describe uncertainty over both business returns and infrastructure spending. Wong’s warning similarly left the path of any correction open, while setting out what he wants Singapore to gain from current demand.

Wong’s call for international AI safeguards

Wong also addressed risks beyond the investment cycle. He said no country could manage alone the possibility of powerful autonomous AI systems going rogue or being misused to cause harm. Rules and standards for frontier AI would ultimately need to be international, he said, suggesting safeguards might be developed under United Nations auspices while acknowledging that such an arrangement would take time.

He said safeguards would also differ by sector. Regulators in healthcare, finance and advanced manufacturing, for example, would need to decide which decisions must remain under human control and what protections fit their activities. As examples already taking shape, Wong cited Monetary Authority of Singapore guidelines on AI in finance and tests of autonomous vehicles, including a Waymo pilot intended to help authorities assess safety requirements before a wider rollout.

For workers, Wong described Singapore’s approach as protecting each worker rather than a particular job. He linked that approach to cooperation among government, employers and unions. His remarks set a direction for using the boom, but the reporting gives no timetable or measure for how many jobs the current demand might create. Nor does it establish that a market correction, an AI loss-of-control incident or a data-centre bust has happened.