Earnings Recap — Q2 FY2026
CY Q3 2026 · Reported July 30, 2026 · Beat 1 of last 6 quarters
The Buildout is a website built on agentic AI tools, currently in beta, and could have factual errors.
Brookfield's accelerating AI infrastructure pipeline, including the Kentucky AI campus (1.2 GW) and South Korea sovereign compute project, underscores the massive scale of capital required for the AI buildout. The expansion of the Bloom Energy framework to $25 billion and the IPO of its data center business highlight the growing role of infrastructure investors in financing behind-the-meter power and compute capacity. The company's ability to originate large-scale, bilateral projects positions it as a key enabler of the AI infrastructure buildout.
Brookfield Infrastructure reported FFO of $702 million, up 10% year-over-year, with strong performance across all segments, particularly Data (+36%) and Midstream (+17%). The company advanced its capital recycling program, generating nearly $1.2 billion in proceeds year-to-date, including the IPO of its U.S. colocation data center business. New investments included the Clarus acquisition in New Zealand and an expanded Bloom Energy framework to $25 billion. Management highlighted accelerating AI infrastructure momentum, with new projects in Kentucky and South Korea, and discussed their growth outlook for 2026.
Management reiterated confidence in delivering 10%+ per unit FFO growth in 2026, supported by organic growth within the 6%-9% target range and contributions from new investments. They expect to achieve their capital recycling objective for the year, with several sale processes underway. On AI infrastructure, they see a massive opportunity, with over 100 GW of load required over the next decade, and expect to deploy significant equity in AI infrastructure, with capital deployment back-end loaded over a 3-5 year horizon. They also plan to complete the corporate simplification to a single publicly traded corporation in Q4 2026, subject to unitholder and shareholder approval on October 14.
“We have definitely seen no reduction in the developments underway or the speed at which our clients are looking to bring forward projects. So while capital markets have obviously pulled back in the last couple of weeks, customers—and our customers are the largest hyperscalers in the world—are obviously thinking about longer-term trends as opposed to short-term gyrations.”
on AI demand resilience
“We think that in excess of 100 gigawatts of load will be required over the next decade. And I guess when you think about that, really hyperscalers are looking for partners who can engage at scale and can really help move the needle from that perspective.”
on AI opportunity scale
“Public markets have been an increasingly effective exit channel for us. So far in 2026, we have generated meaningful proceeds from public market transactions, reflecting both the quality of the businesses we have built and the depth of investor demand for scaled, high-quality infrastructure platforms.”
on Capital recycling
Clearly, the market has become more anxious about the CapEx going into AI and data centers and the timing of the payoff. So I was hoping you could speak to the opportunity set and just how BIP is able to maintain its investment guardrails against that backdrop. And maybe you could touch on whether you are seeing degradation in contract terms more broadly across that space.
Samuel Pollock noted no reduction in development momentum and that customers (largest hyperscalers) think long-term. Lief Williams added that development yields remain high single digits to low double digits, escalators are at the higher end (2.5-3%), and lease terms are extending to 20 years for greenfield projects, indicating strong commercial terms.
It seems like the AI factory strategy is really starting to gain traction here. It is obviously great to see. And it seems like, based on your comments, Samuel, there are still a lot of irons in the fire. Just wondering if you can frame how large of an opportunity the AI factory strategy could be over time, and maybe just for the projects and frameworks that you have secured to date, any thoughts on potential equity commitments or deployment timing from a BIP perspective.
Samuel Pollock said the AI factory strategy could be a significant component of investment over a 3-5 year horizon, but capital deployment is back-end loaded due to development timelines. Lief Williams highlighted the massive opportunity, citing over 100 GW of load required over the next decade, and described the Kentucky project requiring up to $100 billion in private capital.
Notwithstanding all the comments that you just made about NIMBYism and an earlier response about the timing of the payoff for the CapEx in AI, you have obviously been quite successful in signing a number of AI-related deals in Kentucky and South Korea. Maybe we are front-running the Investor Day a little bit, but I am curious whether you see these AI opportunities as progressing in line with your prior projections, or are there pockets of the AI infrastructure value chain that you think may be accelerating?
David Krant said deployment is in line with business plan, with pace of announcements accelerating. Samuel Pollock noted the opportunity set spans AI factories, compute, behind-the-meter power, and adjacencies, with progress across all areas. He highlighted Radiant (neocloud) and Bloom as examples of accelerating opportunities.