Data Center Power Demand Strains Global Energy Infrastructure

The intersection of AI and energy infrastructure is creating a structural demand shock. Siemens Energy's CFO stated explicitly that the company is benefiting from surging AI use driving demand for power-hungry data centers, with visibility extending into the next decade. Mitsubishi Heavy Industries reported that global orders for gas turbines will remain strong due to data center buildout, despite an expected slight decline from 2025 levels. Multiple industrial and utility companies are now framing capex decisions around hyperscaler power commitments, signaling a secular shift in demand patterns.

This is reshaping real estate and utility investment. Data center sites are becoming the most coveted real estate class, driving infrastructure overhauls across North America and Europe. Utilities are filing for emergency Department of Energy loan programs to fund accelerated buildout. Companies like Duke Energy are applying for DOE loans representing "potentially billions of dollars in customer savings," reflecting the scale and urgency of grid modernization. Private credit markets are also mobilizing; investors are increasingly seeking exposure to data center financing, as evidenced by new dedicated real assets intelligence platforms and infrastructure investment vehicles launching this quarter.

Equipment makers are capitalizing. Gas turbine, power distribution, and industrial cooling suppliers are all seeing order acceleration. However, supply chains are tightening; semiconductor and component manufacturers are experiencing competing demand from both AI model training and power infrastructure. Energy security is becoming a geopolitical battleground; nations are racing to ensure stable power access for AI hubs, and capital is concentrating in jurisdictions with lowest-cost, most-stable power.

The debate centers on sustainability. Critics warn that the power demands of AI training and inference at scale may exceed renewable generation capacity in many regions, forcing a return to natural gas and nuclear. Others argue that this capex cycle will ultimately accelerate grid modernization and renewable deployment, with AI itself optimizing energy distribution.