Key Findings
- South Dakota reports the lowest AI-related energy usage per business in the U.S. at 4,314 kWh annually.
- North Carolina and Ohio also rank among the top three, with 4,952 kWh and 5,465 kWh per business respectively.
- Each state keeps AI energy use low, unlike Maine, where consumption more than doubles South Dakota’s.
Not all digital expansion comes with higher energy costs.
Recent DesignRush data shows that some states use up to 40% less power than the U.S. average of 7,274 kWh.
This finding adds a new layer to the conversation on sustainable digital growth: it’s not just about how many companies use automation, but how they implement it.
In contrast to energy-intensive tech hubs, the states of South Dakota, North Carolina, and Ohio maintain relatively lean systems.
They’ve adopted automation carefully, keeping operational overhead low while still integrating modern tools across industries.
Top 3 Energy-Efficient AI States
-content.jpg)
1. South Dakota
- AI Energy Consumption per Business: 4,314 kWh per business
- AI Adoption: 3.55%
South Dakota leads the nation in AI energy efficiency.
With under 2,000 businesses using AI, the state’s total energy impact remains small.
Companies also tend to be smaller, with just 5.7 employees on average, allowing operations to run on lighter systems.
In a move that could reinforce this energy-conscious approach, the state recently approved over $3.3 million for the Big Stone Energy Storage Project, a thermal facility aimed at improving grid stability and energy efficiency.
2. North Carolina
- AI Energy Consumption per Business: 4,952 kWh per business
- AI Adoption: 8.75%
Despite higher adoption than South Dakota, North Carolina keeps energy usage low.
That’s likely due to workforce size, with AI-using businesses averaging just 6.6 employees, indicating streamlined processes and smaller infrastructure needs.
But the state is also being smart about energy planning.
A study from Duke University found that North Carolina could add 1.3 GW of data center capacity (that’s a lot of computing power) without building new gas plants just by using its existing power grid more efficiently.
Also, some data centers in the state are powered entirely by wind energy, like the 208 MW Amazon Wind Farm in eastern North Carolina.
That helps reduce the environmental impact of all that computing.
Combined with business-friendly energy pricing and low utility rates, North Carolina offers a model of how to scale AI adoption while minimizing environmental impact and grid stress.
3. Ohio
- AI Energy Consumption per Business: 5,465 kWh
- AI Adoption: 8.15%
Ohio stands out for its scale. It supports nearly 50,000 AI-active businesses and over 362,000 workers in this space.
Even with those large numbers, energy use per business remains well-controlled, signaling efficient use of shared or virtual computing resources.
Part of this efficiency stems from Ohio’s investment in next-generation infrastructure.
The state is piloting on-site microgrids for data centers, combining renewables, gas, and battery storage.
These systems help mitigate demand spikes and even repurpose waste heat to support greenhouse agriculture, demonstrating a circular approach to tech energy use.
Additionally, utilities like AEP Ohio are reshaping how power costs are distributed.
A proposed structure would require data centers to pay in advance for 75–90% of their projected capacity, helping ensure that grid upgrades don’t fall on regular consumers.
While controversial, this signals a strong focus on grid equity and long-term planning.
Why This Matters
The energy conversation around automation is overdue.
While most headlines focus on growth rates or tech breakthroughs, cost and infrastructure deserve equal attention.
States that keep energy use in check are building tech maturity without overstretching resources.
For leaders, that means lower costs and more predictable operations.
For governments, it’s a sign of an economy that adapts without overloading the grid.
These findings should inform how companies scale and where they choose to expand.
Methodology
This ranking is based on a 2025 dataset analyzing annual AI-related energy usage by business across all 50 U.S. states.
To determine the energy consumption per AI-using business, DesignRush used eight metrics.
Metric | Description | Source |
Total Businesses per State | Number of active businesses in each U.S. state | U.S. Census Bureau (2024) |
AI Adoption Rate (%) | Share of businesses using AI from January–May 2025 | |
Businesses Using AI | Calculated by applying AI adoption rate to total businesses | Derived from Census + BTOS data |
Employees per Business | Average number of employees per business | |
Total Employees Using AI | Estimated from AI-using businesses × average employees | Calculated Metric |
Daily Energy Use per Employee | 2.9 kWh per employee using AI per workday | |
Annual Energy Use (kWh) | Employees using AI × 2.9 kWh × 260 workdays | Calculated Metric |
Energy Use per Business (kWh) | Total annual AI energy divided by businesses using AI | Calculated Metric |
-details.jpg)



-preview.jpg)


