The Next Frontier: AI-Native Infrastructure Funds

The announcement of Project Stargate — the US$500 billion joint venture between OpenAI, SoftBank, and Oracle — crystallised something that had been building quietly for two years: the infrastructure required to power artificial intelligence is now a distinct asset class, demanding dedicated capital structures, specialist operators, and purpose-built fund vehicles. The generalist infrastructure fund, with its mix of airports, toll roads, and utilities, is no longer the right instrument for this moment.

Blackstone's AI infrastructure strategy, Brookfield's data centre platform, and KKR's expanding digital infrastructure portfolio are not rebranding exercises. They reflect a genuine structural shift in how the largest pools of institutional capital are being organised. These managers have concluded — correctly, in my view — that AI infrastructure requires a different underwriting framework, a different operational toolkit, and a different set of relationships than traditional infrastructure investing. The hyperscalers are not passive tenants; they are co-architects of the assets they occupy, and managing that relationship requires deep technical and commercial fluency.

The defining characteristics of an AI-native infrastructure fund are worth examining. First, the asset scope is broader than the data centre alone — it encompasses the power generation and transmission assets that feed compute clusters, the fibre and networking layers that connect them, and the cooling systems that keep them operational. Second, the development risk profile is higher than stabilised infrastructure, requiring fund structures that can accommodate construction exposure and lease-up periods. Third, and most critically, the demand driver is a single technology cycle — which means vintage selection and entry timing matter enormously.

For family offices and sophisticated private investors, the emergence of dedicated AI infrastructure funds creates a genuine portfolio construction question. The return profile — contractual yield plus development upside — sits between traditional infrastructure and private equity. The correlation to public markets is low. The duration is long. For investors who have been underweight real assets, this is a compelling entry point, provided they are selective about managers with genuine operational capability rather than those simply affixing an "AI" label to existing strategies.

At Lensbridge, we have spent the better part of a decade building relationships and insight at the intersection of private equity, digital infrastructure, and technology. The emergence of AI-native fund strategies is the logical culmination of that convergence — and we intend to be thoughtful, disciplined participants in it.