The shift to artificial intelligence in the utility sector is not a passing trend. For Duke Energy, it’s a necessary step to strengthen grid resilience, especially during storm seasons. In 2024, the company’s investment division supported AiDash, a startup using satellite imagery and machine learning to identify tree growth and vegetation that could threaten power lines—well before crews can spot the issues on the ground. The ultimate goal is fewer outages, reduced wildfire risks, and a grid that meets the rising energy demands from data centers and industries.
This shift marks a critical turning point for the grid business. Utilities are evolving from simple delivery systems to data-driven networks managing power load, operational risk, and automated systems that communicate in real time. It’s no longer just about keeping the lights on. Utilities must now invest in AI and cutting-edge technologies to future-proof their grids and ensure reliability in the age of electrification.
Steve Smith, President of National Grid Partners and Chief Strategy & Regulation Officer at National Grid, remarked that AI is both a challenge and an opportunity. He sees AI as potentially the key to solving some of the sector’s toughest problems.
After two decades of relatively stable electricity demand, the U.S. now faces a surge in demand driven by AI-powered data centers, electric vehicle adoption, and the electrification of heating. Smith pointed out that utilities are under increasing pressure to meet these new challenges, especially given the rising expectations for reliable power in a warming world.
National Grid Partners (NGP) has made significant strides since its inception in 2018, investing nearly $550 million in over 50 companies. Seven of these companies have already had successful exits, indicating a strong return on investment.
However, not every venture will succeed. Smith acknowledged that while some investments might not work out, the goal remains clear: bring proven technologies into utility operations and scale them beyond pilot programs.
Real Investments, Real Impact
Across the U.K., National Grid collaborates with Emerald AI, using its “Conductor” software to optimize data center operations. This software adapts to grid conditions, shifting computing loads to reduce power strain. In a demonstration, it reduced peak-period energy usage by 25%.
Meanwhile, AES has partnered with LineVision to install 42 non-contact sensors along transmission lines in Indiana and Ohio. These sensors monitor real-time conditions and apply Dynamic Line Ratings, enabling AES to transmit more electricity safely without upgrading infrastructure. This approach boosts capacity at a fraction of the cost of traditional solutions.
Southern Company, through Alabama and Georgia Power, is testing WeaveGrid’s smart-charging platform. This system automatically schedules EV charging during off-peak hours, promoting electric vehicle adoption without requiring expensive infrastructure or long regulatory approval processes.
Smith explained that National Grid Partners’ strategy is straightforward. The approach is to back companies with proven technologies, integrate them into utility systems, demonstrate their scalability, and then allow them to expand across the industry. Their role is to accelerate growth, making connections quicker and easier by leveraging AI and new technologies.
The Stakes and Rewards of AI in Utilities
Smart grids offer significant advantages to utilities. Faster connections for high-demand sectors like data centers and factories open up new revenue streams. Efficient asset use can delay the need for expensive infrastructure upgrades. Improved forecasting and adaptable load management reduce risks and lower operational costs. Utilities that implement the most advanced grids will be well-positioned in a world focused on cutting carbon emissions.
However, achieving these benefits isn’t a given. The utility sector remains heavily regulated and risk-averse, and investments in disruptive technologies face significant hurdles, including ratepayer concerns, regulatory oversight, and lengthy implementation timelines. The specter of a tech bubble still looms, where hype could outpace practical deployment. As Smith emphasized, not every investment will succeed. Utilities need to remain disciplined in their approach—this isn’t about speculative ventures but strategic investments aimed at operational value.
Despite these challenges, the demand for modernized, resilient grids is undeniable. The tools available are increasingly effective, and the business model for utilities is evolving. They are no longer just technology consumers—they are active partners, investors, and deployment platforms.
Smith’s perspective is optimistic. AI is not just a tool; it’s the means by which utilities can expand capacity, modernize systems, and quickly share knowledge across the industry. Speed is essential, and innovation is now critical for survival.
For utilities, corporate-venture teams, and investors, the message is clear: the era of waiting for innovation to happen is over. Now, utilities must help shape and lead the charge.
