Singapore’s sovereign wealth fund Temasek is making a bold bet on AI, and it’s attached to a target number, not a vague ambition. At its annual review meeting “yesterday,” Temasek said it intends to increase its exposure to AI until it reaches 15 percent in 2031. Right now, the fund has over $400 billion in assets, and around six percent of those are currently tied up in AI companies, including OpenAI.

Temasek also laid out what that actually means in dollars and operating choices. The fund expects its decision could translate into investing at least an additional $36 billion in AI outfits over the next five years. Internally, it’s also “embedding AI into how we invest and operate,” saying it will augment human decision-making, sharpen workflows, and enhance productivity across the firm. In other words, Temasek is not only buying AI exposure. It’s trying to use AI to run the business better, and it’s positioning that as part of why the bet is worth placing.

If you’re an executive trying to interpret this, the key is that sovereign wealth funds are not usually loud about technology roadmaps. Their pitch is patience and performance, not trend-chasing. Temasek, one of the world’s largest investment houses, reported 10.5 percent total shareholder returns last year, and it has delivered 6.8 percent returns over 20 years. The new AI push looks like an attempt to protect that performance record in a world where infrastructure, energy, and compute are the new constraints. The logic is straightforward: if AI demand keeps growing, the winners are not only model builders and cloud vendors, but also the pipes that feed them.

Temasek gave a crisp map of where it wants to place the increased AI exposure. It says it will target five areas: energy and data centres, semiconductors, cloud services providers, foundation models, and AI applications and software infrastructure. That list matters because it spans the full stack. It’s not just “buy the AI company.” It’s also “buy the electricity, buy the chips, buy the cloud capacity, and buy the software layer that businesses will actually run.” The fund also points to why infrastructure is moving up its priority list: it sees rising demand for AI and expects to grow its investment in infrastructure from one percent of its portfolio to five percent.

And Temasek didn’t frame that infrastructure shift as purely about compute. It tied AI growth to electrification and grid modernization, including energy storage and decarbonization. The fund said it sees “compelling opportunities in ageing infrastructure and grid modernisation, renewable and nuclear energy, energy storage, and breakthrough decarbonisation technologies,” underpinned by “rising electrification demand and AI-driven data centre growth.” This is an important second-order signal for decision-makers: AI is treated here as an energy and industrial transformation, not just a software upgrade.

Temasek even named the kind of internal change it wants to make in parallel with investing. It argued that workforce upskilling stays central, saying it invests to accelerate adoption of AI “while keeping people at the heart of workforce upskilling.” It also said it wants AI to keep “people [are] at the centre of this transformation.” CFO and president Ms Png Chin Yee is the specific leader connected to this messaging. For boards and CFOs, the implication is that AI strategy is being packaged as both financial and organizational readiness, not just capex and stock selection.

In market terms, Temasek’s additional spend will not be the only source of AI money. The fund itself acknowledges the reality of hyperscalers. It notes that hyperscale players alone already intend to invest over $1 trillion in AI, so Temasek’s contribution is unlikely to dominate total spending. Yet that doesn’t make the move less consequential. Sovereign wealth fund capital often functions like a confidence signal, especially when it’s tied to a long-term exposure target and anchored by a track record of returns. For investors who worry that AI might not pay off, Temasek’s decision could be used as a pressure-release valve: if a fund built to grow wealth for Singapore is going heavier, it suggests the opportunity set is still expanding.

The operational part of the story is what makes it worth watching beyond headlines. Temasek’s message is that it is “eating its own dog food,” using AI to improve how it invests and operates. That matters because it hints at a feedback loop: invest more in AI, use AI to make better investment and workforce decisions, then allocate more confidently. And historically, Temasek’s tech track record includes Alibaba, OpenAI, and Tencent. It also previously helped Dell to buy storage company EMC in 2016, structuring the deal in a way the fund framed using “the precepts of Chinese numerology.” Whether you focus on the modern AI target or the older tech pattern, the through-line is consistent: Temasek treats technology as strategic, and it plans to keep compounding that view over multi-year horizons.