Utilities form AI consortium to tackle grid challenges while battling power demand

Electric utility companies are forming a new AI consortium to tackle the power sector’s most pressing challenges, including surging electricity demand, grid vulnerabilities, and the clean energy transition. This collaboration between major U.S. power suppliers and tech companies like Microsoft, Oracle, and NVIDIA aims to develop specialized AI models that could transform how energy is generated, distributed, and managed—potentially creating a more efficient, resilient grid while supporting decarbonization efforts, despite the ironic reality that AI itself is a significant driver of increasing power consumption.

The big picture: A new consortium announced Thursday brings together electric utility companies and technology giants to develop AI models specifically designed to address power sector challenges.

• The Electric Power Research Institute (EPRI), an independent nonprofit organization, spearheaded the formation of this consortium that includes most major U.S. power suppliers alongside tech companies like Microsoft, Oracle, and NVIDIA.
• This collaboration will build upon EPRI’s existing work, which already includes a large language model trained on the institute’s data—essentially creating a specialized ChatGPT for utility companies.

Why this matters: The power industry faces a convergence of unprecedented challenges that AI could help solve.

• After years of stable demand, electricity needs are surging both domestically and globally, straining existing infrastructure.
• Increased severe weather events and wildfires are creating more frequent grid disruptions, requiring innovative approaches to resilience.
• The industry must simultaneously decarbonize operations while meeting growing demand, a complex technical challenge.

How AI could help: EPRI’s Executive Director of AI and quantum, Jeremy Renshaw, identified several key applications for artificial intelligence in power systems.

• AI can better synchronize generation forecasting with demand forecasting, potentially reducing reliance on carbon-intensive backup generation.
• The technology could optimize renewable energy deployment by matching production from intermittent sources like wind and solar with peak demand periods.
• AI systems may also accelerate permitting processes for new power sources and transmission lines, addressing infrastructure bottlenecks.

The ironic twist: The technology being deployed to solve power problems is itself a significant contributor to electricity demand growth.

• Some studies project data centers could consume up to 12 percent of all U.S. electricity production by the end of this decade.
• NVIDIA’s Senior Managing Director Marc Spieler acknowledged this paradox but noted the company is working to increase the efficiency of GPU chips used in AI computing operations.

What they’re saying: Industry leaders emphasize that AI can improve multiple aspects of power delivery.

• “What they’re really looking at is how can we have safer, more affordable, more reliable and resilient energy systems,” explained Jeremy Renshaw. “AI can help with all of these.”
• NVIDIA’s Marc Spieler noted the potential of combining industry data: “Together, we can combine the data and train a domain-specific industry model that can help accelerate the industry.”