The next generation of data centers is changing more than how much electricity utilities must supply. It is also changing the speed, concentration and uncertainty of load growth. This demand is arriving amid transmission constraints, lengthy interconnection timelines and competition for grid capacity.
Artificial intelligence is creating large, power-dense loads while also giving utilities tools to automate studies, analyze data and plan faster. That dual role was central to Siemens Grid Software’s webinar, Planning the grid at the speed of AI: From interconnection queues to grid readiness, featuring experts from Siemens, NVIDIA and Dominion Energy. Utilities need more infrastructure, but also smarter, faster and more collaborative planning.
AI factories are changing the scale of demand
Traditional data centers hosted enterprise applications and delivered computing and storage services. AI factories create value through model training and inference. As AI advances toward reasoning models and physical AI systems, its computing and electricity requirements are rising dramatically
According to NVIDIA’s webinar remarks, traditional data center racks typically consume 15 to 25 kilowatts. Current AI infrastructure can require approximately 230 kilowatts per rack, while future systems could approach one megawatt per rack. For utilities, this is not incremental growth. Large loads may cluster geographically and seek service much faster than generation, transmission and distribution infrastructure can be planned, permitted and built.
The grid is part of the AI technology stack
As NVIDIA observed during the webinar, energy has become part of the AI technology stack. AI cannot operate without reliable electricity, making power availability a strategic factor in where AI infrastructure is developed.
Utilities, regulators, system operators, developers, technology providers and customers must coordinate earlier on project certainty, ramp schedules, location, reliability and potential load flexibility.
Dominion Energy’s Northern Virginia experience reinforced that integrating large loads is not solely a utility challenge. Stakeholders must align timelines and assumptions early.
Planning must account for uncertainty
Deterministic forecasts alone become less effective when planners face many possible futures. An AI facility may be delayed, resized, relocated, or phased. Its load profile may change as technology evolves, while renewable output, electrification, distributed resources and transmission constraints add variables.
Utilities should complement deterministic studies with scenario-based and probabilistic planning. Key considerations include project probability and timing, load ramp rates, hourly and seasonal demand, generation and transmission availability, storage performance, flexible-load participation, behind-the-meter resources and permitting timelines.
The goal is not to predict one future perfectly, but to identify decisions that remain sound across many futures.
Flexibility could change the equation
Some AI workloads may be more flexible than commonly assumed. Certain training processes could be shifted, throttled, or scheduled for different times or locations, although inference and other time-sensitive applications may have strict availability requirements. Even partial flexibility could reduce stress during constrained periods, improve use of existing assets and potentially defer selected investments.
Options include workload scheduling, demand response, battery storage and phased energization. Utilities cannot assume flexibility. They need dependable information about which workloads can move, response speed, duration and operational limits.
Behind-the-meter generation may accelerate projects, but it raises fuel, permitting, reliability, market and economic questions.
AI can help utilities plan for AI
Planning teams are evaluating more interconnection requests, operating conditions and investment scenarios, often without equivalent growth in engineering resources. Automation and AI-enabled workflows can expand analytical capacity by handling repeatable tasks, supporting data analysis and helping planners focus on higher-value engineering decisions.
Faster simulation, improved data integration and interoperable platforms can shorten study cycles and create a consistent view of assumptions and results. Combined with probabilistic methods, AI can expand scenario analysis without compromising engineering rigor.
The objective is not to replace engineers, but to give them the speed, scale and information to plan across more possible futures.
Preparing for the AI factory era
Utilities must preserve reliability, safety and affordability while becoming more agile and growth-oriented. Physical reinforcement will remain necessary, but building alone is not enough. Utilities must also use existing capacity more effectively, improve data quality, automate repeatable work and evaluate uncertainty systematically.
AI factories make the urgency visible, but the implications extend beyond data centers. Electrification, renewable integration and distributed resources are increasing grid complexity. AI is creating part of the challenge, but it can also be part of the solution.
Utilities that recognize both sides will be better positioned to support economic growth while maintaining reliable, resilient and affordable service.
Watch Planning the grid at the speed of AI: From interconnection queues to grid readiness on demand to hear experts from Siemens, NVIDIA and Dominion Energy discuss how utilities can prepare for AI-driven load growth.
