Most Data Center Projects Face Failure, Expert Warns
Infrastructure constraints threaten 65% of announced projects over the next 24 months as power demand continues to surge.
Updated on Oct. 27, 2026 in Data Centers
Live Poll
Is now a good time to invest in large-scale AI infrastructure projects?
Investment manager Kevin O'Leary projects that 65% of current data center projects will fail to materialize due to severe infrastructure constraints. These challenges, ranging from power access to political opposition, contrast with soaring demand that is expected to triple compute capacity requirements by Q4 2027.
Why it matters
The massive scale of data center infrastructure, which can cost $16 billion for 1.4 gigawatts of capacity, is colliding with finite regional resources. As projects fight for limited electricity and water, the industry faces a significant correction that could halt the majority of announced capacity expansions.
Large data center builds now require approximately $16 billion for 1.4 gigawatts of capacity. Lawrence Berkeley National Laboratory estimates these facilities could account for 11.8% of U.S. electricity consumption by 2030, within a projected range of 9.5% to 15.3%.
The players
Kevin O'Leary
An investment manager who holds stakes in Bitzero Holdings and oversees private data center development projects in Utah and Alberta.
Lawrence Berkeley National Laboratory
A U.S. Department of Energy national laboratory that conducts research on energy efficiency and national infrastructure trends.
Bitzero Holdings
A data center developer focused on infrastructure projects in regions including Canada, Norway, and Finland.
The details
Data center projects frequently stall due to a mismatch between developer assumptions and physical reality, particularly regarding access to existing natural gas infrastructure for power. Developers often underestimate the hurdles posed by limited fiber connectivity, water scarcity, and state-level moratoriums or local political opposition. These fatal flaws force capital-intensive projects, intended to meet a tripling of compute demand by Q4 2027, to collapse before construction commences.
Timeline
June 2026: Lawrence Berkeley National Laboratory published electricity usage estimates.
October 16, 2026: Kevin O'Leary detailed industry project failure rates.
Next 24 months: Expected window for 60% to 65% of project shutdowns.
Q4 2027: Target date for projected tripling of compute capacity demand.
2030: Year by which data centers are projected to reach 11.8% of U.S. electricity load.
The Tech Race
This contraction follows the Lawrence Berkeley National Laboratory 2026 U.S. electricity consumption report, which underscores the extreme load these facilities place on the grid. O'Leary's assessment suggests that current market growth targets may be physically impossible to reconcile with available power resources.
Investors and industry stakeholders should expect significant volatility in companies relying on announced, but not yet built, data center capacity. The projected shutdown of over 60% of these projects over the next 24 months will likely lead to delays in regional cloud service deployments and increased competition for utility resources.
The takeaway
The data center sector faces a massive consolidation as projects fail to secure necessary power and infrastructure access. Watch for updated project cancellation rates over the next 24 months to see if the actual contraction matches the predicted 65% failure threshold.
Further reading
For broader trends on infrastructure capacity and facility expansion, see Data Centers.
Source note: This article includes information reported by Benzinga.
Live Poll
Is now a good time to invest in large-scale AI infrastructure projects?
