Singapore state investor Temasek aims to increase its AI-related exposure from 6% of portfolio value to up to 15% by 2031, betting on the technology as a structural, long-term driver of value creation. But value capture, a senior Temasek executive said, will come from the other 85%, which must adopt AI to stay competitive.
The AI target excludes exposure held through its Singapore-based Temasek Portfolio Companies (TPCs), and the exposure will be all equity.
Temasek, which views the rapid advancement of AI as a pivotal phase that will create vast new opportunities, has been investing across the AI value chain in five areas: energy and data centres, semiconductors, cloud services providers, foundation models, and AI applications and software infrastructure.
Some portfolio companies are leaders in one area, while others are vertically integrated across several, creating what Temasek Holdings CEO Dilhan Pillay described as a ‘flywheel effect’.
“AI is integral to how we sense emerging opportunities, adapt our portfolio, and thrive as an institution,” said Pillay.
Temasek’s broader AI strategy rests on four pillars: AI-enabling itself, AI-proofing its portfolio, scaling its AI exposure, and supporting AI diffusion across its ecosystem. Internally, it runs a firm-wide Digital Fluency programme, deploys AI agents for specialised, high-value workflows, and is developing an agentic AI strategy.
Across the portfolio, it is actively engaging TPC boards and management teams on AI as a strategic priority, working in partnership with unions and government agencies through tripartism to redesign jobs and reskill, upskill, and uplift the TPC workforce.
The investor has run AI immersion programmes for TPC leaders in Silicon Valley, Shanghai, and Hangzhou.
With AI, Pillay said, the organisational structure of the future will increasingly be built around systems, agents and people—but people must remain at the heart of the transformation. He stressed that AI replaces tasks, not necessarily jobs, adding that it would be incorrect to say there is no role for entry-level talent or young graduates in an AI-driven world, which he described as ripe for them.
“We believe in AI’s promise, but we also take the risks seriously. We want to empower our people to harness AI, not be replaced by it,” he said.
Temasek employs about 1,000 people globally, while its TPCs employ over 160,000 people in Singapore and about 400,000 worldwide.
Chia Song Hwee, Chief Executive Officer of Temasek Global Investment, framed the AI exposure as a natural extension of Temasek’s long-running digitisation investing rather than a portfolio assembled in a short period. Temasek established a multidisciplinary AI pod in 2019 to build expertise in the then-nascent area.
Temasek’s telecommunications, media, and technology (TMT) exposure stands at approximately 23% of the total portfolio and AI will be part of that . In the 2010-12 timeframe, the TMT portfolio was approximately 90% telecommunications, with hardly any technology exposure—a mix that has since changed dramatically, with telecommunications now a much smaller part.
The portfolio is not static and is expected to continue to evolve to reflect what is relevant going forward.
Rohit Sipahimalani, Chief Investment Officer, Temasek International, explained that risk was inherent to the technology’s pace of growth. “Generally speaking, in any breakthrough technology like AI, which is growing at the pace that it is growing, there will be risks,” he said, adding that Temasek manages this in three ways.
The first is scale: even at 15%, he said, the AI exposure would be significant, but still a limited share of the total portfolio. The second is the varied risk profile within the exposure—energy and data centre infrastructure investments, for instance, are largely based on 15-year contracts with hyperscalers, providing contracted cash flows over that period. “That’s much more stable,” he said.
The third is liquidity. Well over 60% of Temasek’s current AI exposure is liquid, Sipahimalani said. “If we see changes in the environment, the ability to pivot is much easier here than it would be in private deals.”
He added that the growth exposure to AI—riskier, but with potentially higher returns—is being balanced by simultaneously increasing allocations to core-plus infrastructure and private credit, more stable areas that would do better in a higher-inflation environment.
“We’re pretty comfortable with the exposure on that basis, recognising there will be ups and downs,” he concluded.
