Berkshire Hathaway Energy Sees AI Data-Center Demand as an Opportunity, With a Retail-Rate Constraint
DataNews Editorial Team
Berkshire Hathaway CEO Greg Abel said power demand from AI data centers represents a significant opportunity for Berkshire Hathaway and Berkshire Hathaway Energy (BHE), while identifying energy scarcity as a constraint on AI expansion. He said BHE would not sell power to hyperscalers if doing so raised rates for other customers.
- BHE identifies AI data-center load as a significant opportunity, according to Greg Abel.
- Energy scarcity was cited as a constraint on AI development.
- BHE would not serve hyperscaler demand if it increased rates for other customers.
- The source discloses no project, contract, capacity volume, investment, or timetable tied to AI data centers.
- BHE reported $891 million in second-quarter 2026 operating income, up 27% year over year.
Greg Abel, chief executive of Berkshire Hathaway, has positioned AI data-center electricity demand as a significant commercial opportunity for Berkshire Hathaway and Berkshire Hathaway Energy, while drawing a clear boundary around customer-rate impacts.
In an interview with CNBC, Abel said energy scarcity is constraining AI development. He described data centers as a significant opportunity for Berkshire Hathaway and BHE, but said the utility business would not sell electricity to hyperscalers if that resulted in higher rates for other customers.
The stance matters for large-load development because it links hyperscaler supply arrangements to the cost consequences for incumbent customers. BHE's position suggests that the commercial viability of new AI load may depend not only on available generation and grid capacity, but also on whether the costs of serving that load can be isolated or otherwise addressed without shifting costs to the broader customer base. This is an inference from Abel's stated position; the source does not describe any specific mechanism BHE is using in practice.
BHE reported operating income of $891 million in the second quarter of 2026, up 27% year over year. The reported figures provide financial context for the energy business, but the source does not connect them to a specific AI data-center transaction or development.
No particular hyperscaler, data-center site, service territory, power volume, interconnection arrangement, investment amount, contract, construction activity, or delivery schedule was disclosed. Abel's remarks therefore represent a public view of market opportunity and a condition for serving demand, rather than an announced project or confirmed power-supply agreement.
The comments signal that a major energy business views AI load as a potential growth driver, but only where new supply does not worsen the economics of service for existing customers. For hyperscalers and data-center operators, this reinforces that available power and the allocation of grid and tariff impacts remain central to interconnection and procurement decisions.
- — / September 24, 2026 · EnglishSource →
This article was prepared by the DataNews editorial team based on the sources listed above.