Thought leadership on digital infrastructure, artificial intelligence, and private capital — by LN Sadani, CEO, Lensbridge Capital.
GIC, Temasek, Khazanah, and ADIA are quietly reshaping the digital infrastructure landscape across Asia. What does their growing appetite for AI-linked assets mean for private investors?
Blackstone, Brookfield, and KKR are raising dedicated AI infrastructure vehicles. This is not a rebranding exercise — it represents a fundamental restructuring of how institutional capital flows into the digital economy.
From Stargate to sovereign AI mandates, nuclear power deals to data centre records — 2025 was the year the AI infrastructure thesis moved from conviction to consensus.
The AI investment cycle is maturing. Investors who deployed capital in 2021–2023 are beginning to ask the question that always comes after the hype: where are the returns?
SoftBank, NTT, and the Japanese government are positioning Japan as Asia's premier AI compute hub. For investors who have overlooked Japan's digital infrastructure story, the window is reopening.
As AI becomes embedded in portfolio companies and investment processes alike, the governance frameworks that LPs and GPs rely on are struggling to keep pace. This is a risk that is not yet fully priced.
Not all AI fund vintages are created equal. The secondaries market is emerging as the most efficient mechanism for investors to reposition their AI exposure — and for GPs to manage legacy portfolios.
Johor, Jakarta, and Chennai are no longer overflow markets. They are becoming primary destinations for hyperscaler investment as Singapore, Hong Kong, and Tokyo hit capacity and cost constraints.
As banks tighten lending standards for data centre construction, private credit funds are stepping into the gap — offering flexible, bespoke financing that traditional lenders cannot match.
Agentic AI — systems that plan, reason, and act autonomously over extended periods — is generating a new class of infrastructure requirements that the current data centre buildout was not designed to meet.
The Asian data centre market is entering a consolidation phase. Smaller operators face a stark choice: scale up, find a strategic partner, or be acquired. For investors, this creates a well-defined opportunity set.
Microsoft's deal to restart Three Mile Island was the headline, but the real story is structural: data centres need firm, carbon-free power at scale, and nuclear is the only technology that can deliver it.
The release of DeepSeek R1 sent Nvidia's stock down 17% in a single day. But the real question for infrastructure investors is not whether efficient AI reduces demand — it is whether it changes the nature of demand.
The announcement of Project Stargate — a US$500 billion AI infrastructure joint venture backed by the US government — is either the most important infrastructure commitment of the decade or the most ambitious piece of industrial theatre. Possibly both.
Hyperscalers spent over US$200 billion on capital expenditure in 2024. Sovereign wealth funds entered the data centre market. Nuclear power became a serious topic in boardrooms. This was the year AI infrastructure went mainstream.
The infrastructure of the internet — cell towers, subsea cables, and fibre networks — is being repriced as AI demand transforms the economics of connectivity. For real asset investors, this is a moment of genuine opportunity.
Nvidia has captured the imagination of investors, but the AI supply chain extends far beyond GPU manufacturers. The picks-and-shovels opportunity in AI infrastructure is broader and deeper than most investors realise.
After a decade of cloud migration, enterprises are bringing workloads back on-premises. The hybrid infrastructure model is reshaping the data centre market — and creating new investment opportunities in the process.
GP-led continuation vehicles have moved from a niche liquidity tool to a mainstream exit mechanism. For investors who understand how to evaluate them, they represent one of the most interesting opportunities in private markets today.
AI is simultaneously the most powerful tool available to cybersecurity defenders and the most dangerous weapon in the hands of attackers. The investment implications of this duality are significant.
Family offices are among the most natural owners of digital infrastructure assets. Long investment horizons, flexible capital structures, and the ability to take concentrated positions make them ideally suited to this asset class.
India is building the digital infrastructure of a US$5 trillion economy. The scale of the opportunity — and the speed of execution — is unlike anything else in the world today.
The AI compute buildout is generating enormous demand for programmable, high-performance networking. Network-as-a-Service operators are the unsung heroes of the AI infrastructure stack.
AI data centres consume millions of litres of water for cooling. As water scarcity becomes a global challenge, the water footprint of AI infrastructure is emerging as a material investment risk.
From the UAE's Falcon to France's Mistral, governments and national champions are investing in homegrown AI models. Sovereign AI is not just a technology story — it is a geopolitical and investment story.
Microsoft, Google, Amazon, and Meta are in an arms race for AI infrastructure. The capital expenditure commitments being made today will shape the digital economy for the next decade — and create significant opportunities for private investors.
From ChatGPT's mainstream breakthrough to the OpenAI board crisis, 2023 was a year that tested every assumption about AI investing. Here is what we learned.
Sam Altman was fired and rehired in five days. The OpenAI board crisis exposed the governance vulnerabilities of AI companies — and the risks they pose for investors who have not done their homework.
Institutional investors are reclassifying data centres, fibre networks, and cell towers alongside airports and toll roads. The implications for portfolio construction — and for asset pricing — are significant.
The release of Llama 2 as a free, open-source model changed the competitive dynamics of the AI industry overnight. For investors, the open-source moment raises fundamental questions about where value accrues in the AI stack.
Rising interest rates have frozen the primary private equity market. For secondary investors with dry powder and sector expertise, this is one of the most attractive entry points in a decade.
US export controls on advanced chips to China, the CHIPS Act, and Taiwan's strategic importance have made semiconductors the most geopolitically charged technology in the world. For investors, this creates both risk and opportunity.
Singapore's moratorium on new data centres has been lifted. Hyperscalers are rushing into Malaysia, Indonesia, and Thailand. The Asian data centre market is entering its most dynamic phase yet.
Training GPT-4 reportedly consumed as much electricity as 1,000 US homes use in a year. As AI models grow larger and more widely deployed, the energy implications are becoming impossible to ignore.
Nvidia H100 waitlists stretching to six months. Cloud providers rationing GPU access. The compute scarcity of 2023 is reshaping the AI industry — and creating investment opportunities that most investors have not yet identified.
GPT-4 launched on 14 March 2023 and changed the conversation about AI forever. For infrastructure investors, the question is no longer whether AI will drive demand — it is how fast and how large.
Two months after ChatGPT crossed one million users in five days, the investment community is still catching up. Here is a framework for thinking about what generative AI means for private capital allocators.
I am launching a new thought leadership series — The AGI-le Investor — to share my perspective on the investment themes I believe will define the next decade: artificial intelligence, digital infrastructure, and the private capital strategies that connect them.