Blackstone says data centers remain attractive investments, despite AI rush

Jon Gray, president and chief operating officer of Blackstone, told investors at the Goldman Sachs Financial Services Conference in New York on December 10, 2025 that data centers continue to offer compelling investment opportunities despite a surge of capital into AI hardware and infrastructure. Gray pointed to power constraints and long lead times as structural reasons why high quality, well leased data center assets remain scarce and sought after. He described Blackstone’s cautious underwriting approach, generally securing 15 year or longer leases with major corporate tenants before breaking ground.

Gray referenced McKinsey estimates that he said show a global data center buildout that could total trillions of dollars by 2030, a figure Reuters reported. His comments underscore continuing institutional appetite for physical AI infrastructure even as competition intensifies and financing pours into the sector.

The investment logic carries ramifications beyond balance sheets. Data centers are unusually demanding customers for electricity and water, and their siting often intersects with communities already vulnerable to infrastructure shortfalls. Utilities must plan and pay for upgrades to transmission lines and substations to serve hyperscale facilities, work that can take years and imposes costs on ratepayers or local governments. That dynamic helps explain why Gray emphasized long leases and careful underwriting, but it also means decisions by investors can lock communities into decades of land use and energy commitments.

Public health officials say the stakes are significant. Hospitals, emergency services, and public health laboratories depend on reliable electricity and on local grids that can withstand growing loads. A rapid buildout of data capacity without coordinated grid planning could increase the risk of outages or force utilities to rely more heavily on diesel backup generators, with implications for air quality and respiratory health in adjacent neighborhoods.

Environmental justice advocates note that data centers have frequently clustered in rural counties and in economically distressed suburban areas where land is available and permitting is faster. Those locations can gain jobs and tax revenue, but the benefits are uneven. Long leases and large corporate tax agreements may not translate into sustained local hiring or durable community investment. Water intensive cooling systems can strain local supplies in drought prone regions, compounding existing inequalities in access to clean water.

Policymakers face a twofold challenge. They must accommodate the projected growth in computing capacity while protecting public health and ensuring that local communities share in the economic benefits. That will require tighter coordination among utilities, regulators, developers, and local governments, clearer rules for equitable siting, community benefit agreements that guarantee local hiring and reinvestment, and expedited but transparent permitting that assesses cumulative environmental impacts.

Blackstone’s insistence on securing long leases with major corporate tenants reduces certain financial risks for investors, but it does not resolve broader social trade offs. As trillions of dollars in capacity are planned and financed, decisions made by private equity and corporate tenants will shape energy priorities, land use, and public health outcomes for decades. Regulators and communities will need to push for resilience, equity, and transparency as the AI driven expansion of digital infrastructure accelerates.