Brookfield (NYSE: BAM) unveils Radiant cloud business focused on leasing AI chips globally

Brookfield Asset Management Ltd. (NYSE: BAM; TSX: BAM) is reportedly launching a $10 billion artificial intelligence infrastructure fund alongside a new cloud business that will lease AI chips directly to enterprises and developers, according to reporting by The Information and Reuters. The initiative signals Brookfield’s formal entry into the AI infrastructure supply chain, combining its global data center footprint with capital formation in response to surging demand for high-performance compute (HPC) capacity.

The move comes as capital markets begin bifurcating between hyperscaler incumbents and integrated asset managers pursuing sovereign-backed infrastructure strategies. Brookfield’s new platform, reportedly named Radiant, will operate at the intersection of physical infrastructure, compute access, and leasing models, positioning it to compete with both traditional cloud vendors and AI-native hardware startups.

Why is Brookfield deploying an AI chip-leasing cloud business when hyperscalers already dominate?

Brookfield’s entry into the chip-leasing market through its Radiant cloud platform is not a late move into hyperscale, but a distinctly different proposition. Unlike Amazon Web Services or Microsoft Azure, which build and operate cloud infrastructure tied to software ecosystems, Brookfield is entering as a physical asset allocator. It is leveraging control over land, power, and data center real estate — domains where hyperscalers often face bottlenecks or require partners.

Radiant’s model reportedly focuses on leasing access to AI chips housed in data centers directly to AI developers, offering compute capacity as a standalone service unbundled from proprietary software stacks. This leasing strategy resembles what CoreWeave and Lambda Labs have built, but Brookfield’s scale and real asset integration may allow it to offer more favorable economics and security guarantees to sovereign or institutional buyers.

Brookfield has previously disclosed plans to deploy $100 billion in AI-related infrastructure globally, and the new $10 billion AI fund appears to be an equity vehicle within that framework. Partner geographies reportedly include Qatar, Sweden, and France, which are countries with strong data sovereignty agendas and national AI strategies, and expected to provide Radiant with a strategic diplomatic overlay.

How does Brookfield’s approach differ from Nvidia-backed CoreWeave and others in the AI leasing space?

While CoreWeave has captured headlines through its rapid rise as an Nvidia-backed hyperscaler alternative, Brookfield is pursuing a de-risked, infrastructure-first model rather than a vertically integrated compute startup. Where CoreWeave is reliant on public markets and aggressive hardware purchases, Brookfield can build leased compute capacity on top of existing capital partnerships, including its multi-billion-dollar funds, sovereign investors like the Kuwait Investment Authority, and institutional limited partners with long-term infrastructure mandates.

Moreover, Brookfield is already a dominant force in data center development through its ownership of Compass Datacenters and significant power infrastructure assets. Radiant’s access to these vertically integrated platforms means lower cost per kilowatt, faster deployment cycles, and better pricing control.

The strategic difference is also in customer targeting. Radiant will likely focus on large institutional buyers, including sovereigns, universities, regulated industries, who prefer neutral, non-U.S. hyperscaler-aligned infrastructure for compute access, particularly in AI workloads involving healthcare, defense, or national datasets.

Could this signal the rise of AI infrastructure funds as a new private equity asset class?

Brookfield’s $10 billion fund is being structured as a dedicated vehicle to finance AI infrastructure deployment, reportedly through long-term lease agreements and co-located real estate-financed data center assets. If successful, this could mark a shift in how AI infrastructure is financed, moving away from tech equity and toward private infrastructure capital pools, a space historically dominated by toll roads, utilities, and airports.

AI infrastructure funds could offer stable, inflation-linked returns by monetizing the compute shortage through medium- to long-term leasing contracts. Given Brookfield’s historical expertise in cash-flow-generating real assets, its entry validates this asset class for other infrastructure GPs such as Blackstone, Macquarie, and KKR — all of whom are exploring AI-adjacent infrastructure themes like edge data centers, green power, and submarine cable networks.

This model also allows sovereign wealth funds and pension capital to deploy into AI without taking direct exposure to volatile GPU cycles or unprofitable startups. The chip-leasing layer provides annuity-like economics if properly underwritten, especially when linked to regulated power and land access.

What are the early execution risks and where could Radiant fall short?

Brookfield’s Radiant model, while compelling, faces several non-trivial risks. First, chip supply remains constrained globally, and even Brookfield’s balance sheet may not guarantee priority access to Nvidia H100s or forthcoming B100s. CoreWeave, Amazon Web Services, and Microsoft already have long-term allocations locked in, and the leasing model hinges on chip availability.

Second, market timing risk is acute. Generative AI workloads are still evolving rapidly, and leased compute demand could shift from GPUs to ASICs or optoelectronic chips in under five years. Brookfield’s data center capital cycles are longer and optimized for steady returns, not rapid tech refreshes.

Third, software remains the moat in most AI cloud value chains. By offering raw compute rather than integrated developer services, Radiant could struggle to attract startups and mid-tier users who prefer platform-level abstraction. Without its own SDKs or orchestration layer, Radiant must remain focused on scale buyers who want bare metal access — a narrower market.

Finally, Brookfield will need to navigate data sovereignty regulations in each market it operates. Leasing compute tied to national datasets may trigger compliance challenges unless Radiant guarantees on-premise hosting, local ownership vehicles, or trusted vendor status.

How are investors reacting and what does this mean for hyperscale competition?

Brookfield Asset Management shares did not show immediate volatility on the news, but investor sentiment appears to be warming toward infrastructure-led AI plays as alternatives to frothy chip and software names. The company’s prior announcement of a broader $100 billion AI infrastructure push was met with favorable commentary from institutional analysts, especially those covering real assets, green infrastructure, and sovereign mandates.

For the broader market, Brookfield’s entry into compute leasing is another signal that cloud infrastructure is becoming fragmented. Where Amazon Web Services and Microsoft Azure dominated the first wave of enterprise cloud, the next wave of AI workloads appears to be opening space for asset managers, sovereign entities, and domain-specific clouds like CoreWeave, Lambda, and now Radiant.

Radiant’s success would not only diversify Brookfield’s earnings profile but could reshape capital allocation trends across infrastructure private equity, setting a new benchmark for what qualifies as core-plus infrastructure in the AI era.

Key takeaways: Brookfield’s Radiant platform marks a structural shift in AI infrastructure investing

• Brookfield Asset Management is launching a $10 billion AI infrastructure fund and a new cloud business named Radiant focused on leasing AI chips directly to developers.
• Radiant leverages Brookfield’s existing data center, land, and power assets, allowing vertically integrated deployment of AI compute at global scale.
• The initiative targets sovereign and institutional clients looking for trusted, hyperscaler-independent access to compute resources amid a global GPU shortage.
• Brookfield’s fund structure could validate AI infrastructure as a distinct private equity asset class, offering stable cash flows via long-term leasing models.
• Radiant will compete with CoreWeave, Lambda Labs, and select hyperscalers, but focuses more on neutral infrastructure than software stack bundling.
• Execution risks include chip availability, fast-evolving workload architectures, regulatory compliance, and the absence of a developer-facing software layer.
• Early investor sentiment is cautiously optimistic, with Brookfield’s broader AI infrastructure roadmap seen as a potential secular growth lever.
• The rise of Radiant could pressure hyperscalers by enabling sovereign compute strategies and fragmenting cloud concentration in critical markets.