Stock Markets February 24, 2026

Brookfield Acquires Ori Industries and Rolls It Into Radiant to Rent AI Chips

Deal folds Saudi Aramco-backed Ori into a new unit offering on-demand GPU access for sovereign clouds, with long-term rental contracts to limit technology risk

By Hana Yamamoto BAM
Brookfield Acquires Ori Industries and Rolls It Into Radiant to Rent AI Chips
BAM

Brookfield Asset Management has purchased Ori Industries and merged it into Radiant, a newly formed company that will lease artificial intelligence chips on demand. Financial terms were not disclosed. Radiant will target companies and governments building sovereign clouds and will use long-term rental contracts to lock in revenue across the expected five-year useful life of GPUs to shield against value declines and obsolescence.

Key Points

  • Brookfield acquired Ori Industries and merged it into Radiant, a new company offering on-demand AI chips; financial terms were not disclosed.
  • Radiant will target companies and governments building sovereign clouds, where data is kept within national borders.
  • Brookfield plans to lock rentals across an estimated five-year chip life and expects customers to be investment grade, passing technology risk to lessees.

Brookfield Asset Management on Tuesday completed the acquisition of cloud-computing firm Ori Industries and folded the business into Radiant, a new entity created to offer on-demand access to chips used for artificial intelligence workloads. The company did not disclose the financial terms of the purchase. Ori Industries had previously received backing from Saudi Aramco’s venture arm.

Radiant will market its service to both corporate and government clients that are constructing so-called sovereign clouds - highly controlled computing environments designed so that data does not leave national borders. The offering is intended to address customer demand for lower costs and to shorten long wait times for chips.

To protect against declines in chip value, Brookfield is putting in place contracts that lock rental payments over what it estimates to be a five-year life for each chip. Under those contracts, customers - who Brookfield expects to be investment grade - would remain responsible for full payments even if they cease to require the hardware, according to comments from company executives.

"We will not be taking any technology risk in this business," Sikander Rashid, Brookfield’s head of AI infrastructure, said. That stance reflects a broader industry concern: other infrastructure investors have largely steered clear of major commitments to GPU-as-a-service businesses because of worries that chips can lose value or become obsolete.

Brookfield has been expanding its footprint in AI infrastructure after prior investments in power utilities and data centers. The firm estimates artificial intelligence will require $7 trillion in capital investment overall, with roughly $3 trillion directed to computing infrastructure. "I think of it as a leasing business," Rashid added. "Hopefully we are the first of many who will unlock large-scale capital to the theme."


Context and market impact

  • Radiant’s focus on sovereign clouds positions the unit at the intersection of cloud services, government procurement, and data-center infrastructure.
  • Long-duration rental contracts are designed to shift technological obsolescence risk away from the owner and onto contracted customers, assuming those customers meet expected credit standards.
  • Brookfield’s move underscores increased investor interest in the financial structuring of AI hardware exposure rather than direct technology development.

Risks

  • Chips may lose value or become obsolete, a central concern that has caused other infrastructure investors to avoid large GPU-as-a-service commitments - affects semiconductor and data-center sectors.
  • Counterparty credit risk if contracted customers fail to meet payments despite multi-year rental obligations - affects lending, leasing, and data-center operators.
  • Execution and capital scale risk tied to Brookfield’s estimate that AI requires $7 trillion in capital, including $3 trillion for computing infrastructure; large capital mobilization is required to realize the opportunity - affects infrastructure and capital markets.

More from Stock Markets

Thomson Reuters Shares Jump After CoCounsel Reaches One Million Users Feb 24, 2026 DA Davidson Initiates Rubrik at Buy, Sees Pullback as Attractive Entry Point Feb 24, 2026 Transport Canada Clears Remaining Gulfstream Models After U.S. Tariff Threat Feb 24, 2026 AMD and IBM Drive Mega-Cap Strength as Market Shows Wide Divergence Feb 24, 2026 Istanbul bourses slip as Sports, Transport and Textile & Leather weigh on BIST 100 Feb 24, 2026