July 9, 2026, (Inside AI) — Asian investors are recalibrating their artificial intelligence strategies, shifting focus from headline-making AI stocks to companies that can both benefit from and withstand AI-driven disruption. The cautious pivot was a central theme at the Reuters NEXT Asia event in Singapore, where top fund managers detailed how they are navigating a market increasingly wary of frothy valuations and uncertain returns on massive infrastructure spending.
Rohit Sipahimalani, chief investment officer at Singapore state investor Temasek, captured the dual mandate. Temasek, which holds stakes in Anthropic and OpenAI, plans to boost AI exposure from 6% to as much as 15% over five years. Yet Sipahimalani stressed resilience, noting a shift toward hard assets less vulnerable to AI upheaval.
“You want to ride that trend,” Sipahimalani said. “But the equally big issue is disruption because of AI to many other businesses… We’ve increased our exposure to businesses that are more around hard assets, which are likely to be less disrupted by AI.”
The remarks reflect a broader industry realization: the AI boom’s downstream effects could upend traditional business models, making defensive positioning crucial. Temasek’s approach mirrors a growing preference for “picks and shovels” plays—firms providing essential infrastructure rather than betting on unproven applications.
Read: AI Investors May Pivot to Hyperscalers from Chipmakers, Says Morgan Stanley
Stephanie Hui, head of private and growth equity Asia-Pacific at Goldman Sachs Asset Management, echoed the sentiment, admitting it’s too early to pick winning AI applications. Her firm has invested in a liquid cooling specialist and data centers—the physical backbone of AI expansion.
“I am not smart enough to tell you today which applications are going to be winning, it’s way too early,” Hui said. “We are not going for the front end at this moment… We are going for the simple stuff that facilitates an end proxy for AI adoption.”
This infrastructure-first strategy is gaining traction as global markets, powered by AI euphoria, hit record highs. The Nasdaq Composite has surged over 30% in the past year, driven by chipmakers like Nvidia. But the rally has sparked bubble fears, with the Philadelphia Semiconductor Index experiencing multiple 5% single-day drops in recent months. Investors recall the dot-com bust, where excessive capital flowed into unproven internet ventures with little regard for fundamentals.
Fred Hu, chairman of China’s Primavera Capital Group, voiced caution, questioning the sheer scale of AI spending. “I’m a big believer in the AI revolution but as valuations keep going up, as more and more capital goes into AI… it begs the question, how much is enough,” he said.
Satoshi Ueyama of Bain Capital Japan highlighted the need for end-user demand to justify infrastructure investments, noting his firm targets AI-enabled winners in services and consumer applications. “AI is real but at the same time there’s no denying some parts of the markets are over-excited… Not all AI investment is going to be successful at this stage,” he said.
Underpinning the shift is a hunt for tangible cash flows. Sipahimalani distinguished between areas of “froth” and those with “real cash flows,” emphasizing a value-chain-wide approach. This pragmatism is echoed by a recent Goldman Sachs report noting that AI infrastructure spending could reach $1 trillion in coming years, yet the revenue payoff remains uncertain. By contrast, data center demand is projected to grow 15% annually, offering more predictable returns.
Read: Investors Position for Slower AI Spending Growth by Hyperscalers
The Asian pivot also reflects regional dynamics. Singapore, a hub for data centers and semiconductor manufacturing, is seeing increased investment in cooling technologies and power management—critical for energy-hungry AI systems. China’s tech crackdown has made investors wary of consumer-facing AI apps, pushing capital toward industrial AI and robotics. Japan’s aging population drives demand for AI in healthcare and automation, aligning with Bain’s focus on services.
Historical parallels loom large. The telecom bubble of the late 1990s saw trillions wasted on fiber-optic networks that took years to monetize. Today’s AI infrastructure buildout—from GPU clusters to subsea cables—risks similar overcapacity if end-user adoption lags. Yet proponents argue AI’s utility across sectors makes it fundamentally different from past speculative frenzies.
As the Reuters event made clear, the smart money in Asia is hedging bets—embracing AI’s potential while fortifying against its disruptions. The strategy may define the next phase of the AI investment cycle.
