Congress Holds Hearing on Energy and AI Data Centers — AI: The Washington Report

• The emergence of DeepSeek has complicated the outlook for the amount of investment necessary in data centers and electricity generation for AI advancements. Against that backdrop, on March 5, the House Committee on Energy and Commerce’s Subcommittee on Energy held a hearing titled “Scaling For Growth: Meeting The Demand For Reliable, Affordable Electricity.”
• At the hearing, four experts testified about the growing demand for energy in the US, which they attribute partly to the rise in AI data centers. Representatives from both parties inquired about energy demands for data centers, but lawmakers did not indicate that any particular concrete legislation is forthcoming to increase the US energy supply or build AI data centers.
• The hearing comes as the Trump administration aims to promote private investments in domestic data centers and AI infrastructure to stimulate AI innovation. Just this week, Taiwan’s largest chip manufacturer announced a $100 billion investment in chip manufacturing plants in the US.
• Lawmakers from both parties appeared supportive of increasing the US energy supply to power AI data centers. However, GOP lawmakers are expected to wait for the Trump administration to reveal its AI Action Plan in mid-July before they pursue significant action on AI and data centers. 
   

 
So far this year, the Trump administration’s approach to promoting American AI dominance has involved investing in domestic data centers and increasing electricity generation. As we previously covered, the administration announced $500 billion in private investments for new data centers while also pledging to expedite permitting to accelerate their development. But the January emergence of DeepSeek technology, a Chinese AI model that rivals US models but operates on less powerful chips and consumes less energy, created uncertainty about the administration’s approach by demonstrating that not all AI advancements may necessarily require large energy supplies.

However, in a House Committee on Energy and Commerce hearing on March 5, members of Congress heard from experts in industry and academia who beat the drum for investments in domestic data centers and increased energy production as a way to secure US AI dominance. And while lawmakers from both parties appeared supportive of increasing the US energy supply to power AI data centers, below we discuss key takeaways from the hearing and why lawmakers will likely hold off from acting on AI anytime soon.

Committee Hearing: Opening Remarks

On March 5, the House Committee on Energy and Commerce’s Subcommittee on Energy held a hearing titled “Scaling For Growth: Meeting The Demand For Reliable, Affordable Electricity.” The hearing explored “the challenges faced by utilities and grid operators to deliver reliable, affordable electricity to meet the growing demand for power across the nation.”

In his opening remarks, subcommittee Chair Bob Latta (R-OH) attributed “historic increases in electricity demand” to “energy intensive AI models and domestic manufacturing.” The Chair asserted that the current demand for energy exposes “key impediments to the ability of utilities, grid operators, and generators to keep the lights on” and supply enough energy for technological advancements.

The “stakes could not be higher” for the US to keep up with its energy demands, according to the Chair. “Nations like communist China, who does not share our democratic values,” remarked Rep. Latta, “are seeking to develop world leading AI models through an authoritarian, military lens to export their command-and-control style of governance across the world.” While Rep. Latta views the present moment as an opportunity for the US to “correct course,” “to get there, we need more energy and we need it fast.”

Industry and Academic Testimony

The Committee received testimony from the following witnesses at the hearing: 

• Todd Brickhouse, CEO and General Manager of an electric power cooperative
• Asim Z. Haque, Senior Vice President for Governmental and Member Services of an electric utility company
• Noel W. Black, Senior VP of Regulatory Affairs of an electric utility company
• Tyler H. Norris, James B. Duke Fellow, Duke University

 
The witnesses’ testimonies and responses to the representatives’ questions focused on four main themes: 
 

• Increased Energy Demand is Linked to Data Centers for AI 
   
  • The witnesses, as well as several representatives, attributed the increased energy demand in the US to new data centers for AI and other emerging technologies. Mr. Black said that demand growth is “driven by data centers and artificial intelligence.” While Mr. Norris acknowledged that the release of DeepSeek creates “uncertainty” about how much energy is needed for AI advancements, he forecast that “data centers are expected to account for the single largest growth segment” in the energy space, adding “up to 44% of US electricity load growth through 2028.” 
     
• Ceding AI Dominance to China 
   
  • Several Republican representatives claimed that the US is falling behind China in the AI race because the US does not have adequate energy production capabilities to support AI data centers. “My concern today is we’re in an arms race for artificial intelligence,” stated Rep. Gary Palmer (R-AL). “And it’s clear we don’t have the capacity to build the data systems that we need in order to win that arms race with China.” Rep. Brett Guthrie (R-KY) echoed similar concerns, arguing that if the US doesn’t act on energy production, “we’re going to lose the battle for AI to China.” 
     
• The Need for Permitting Reform and Streamlined Interconnection Procedures 
   
  • All of the witnesses spoke about delays with the permitting process and the interconnection queue and the need for permitting reform to ensure that the US can meet its energy demands. Mr. Norris testified that load growth is “colliding with barriers to timely resource expansion, including supply chain constraints, restrictive intervention procedures, and extended permitting processes.” Mr. Brickhouse called on Congress to improve the permitting process and cut “costly and burdensome regulations to accelerate deployment and maintenance of electric infrastructure.” While several representatives from both parties inquired about permitting and interconnection delays and appeared supportive of reforms, they did not mention any forthcoming proposed substantive legislation to address these issues. 
     
• Load Flexibility for AI Data Centers 
   
  • Mr. Norris spoke at length about flexible load operations for AI data centers, which may provide a possible path to providing such centers with sufficient energy. “Load flexibility,” according to Mr. Norris, “refers to the ability of end-use customers to temporarily reduce their electricity consumption from the grid during periods of system stress by using on-site generators, shifting workload to other facilities, or scaling back operations.” Mr. Norris projected that “if new AI data centers can adjust their electricity consumption during a limited number of hours when power grids experience peak stress, equivalent to 0.5% of their maximum uptime, the existing US power system could accommodate up to four or five Project Stargates,” referring to the $500 billion US investment in AI data centers that we’ve written about. 
     

We do not believe that this hearing will likely lead to the imminent introduction of legislation; it does make apparent that DeepSeek has not altered the appetite and clarion call for increasing the US energy supply to power AI data centers. Thus far, however, neither party has put forth a policy roadmap for AI in this Congress, and the current AI bills introduced in Congress do not propose a comprehensive regulatory approach or touch on data centers and energy infrastructure. As we’ve written about, the Trump administration’s AI action plan, which is currently scheduled to be announced by mid-July, is expected to be the bellwether for the US’s approach to AI, and GOP lawmakers are likely to look to the plan to shape their AI priorities.

We will continue to monitor, analyze, and issue reports on developments about the 119th Congress’ approach to and policies for AI.

Subscribe To Viewpoints