Streamlining AI's Energy Needs for a Sustainable Future

On February 26, 2026, key figures from the AI industry, including Microsoft President Brad Smith (https://www.microsoft.com/en-us) and OpenAI CEO Sam Altman (https://www.openai.com) , will be addressing U.S. lawmakers to advocate for important changes in legislation regarding artificial intelligence. The focus is on the urgent need to streamline federal permitting processes for energy resources and enhance access to government data sets that could significantly boost AI training.

Brad Smith's written testimony for the Senate Commerce Committee's hearing on "Winning the AI Race" emphasizes that America’s infrastructure is outdated, unable to meet the surging electricity demands driven by AI advancements, reshoring manufacturing, and broader electrification trends. He argues that without modernized infrastructure, the U.S. risks falling behind in the global AI race.

Sam Altman will highlight the growing appetite for AI systems, stating that as these technologies improve, their demand will escalate. He notes that meeting this demand requires an increase in essential resources—chips, training data, energy, and supercomputers. Altman envisions a future where AI acts as a 'brain for the world,' accessible to everyone while maintaining safety through sensible restrictions.

Adding to the discussion, CoreWeave CEO Michael Intrator (https://www.coreweave.com) points out the astounding energy requirements of AI computation. He cites estimates from the Energy Department predicting that data centers could consume 12% of U.S. electricity by 2028, up from 4.4% in 2023. Intrator's testimony underscores the urgent need for efficient permitting processes to facilitate the establishment of new energy generation sources and the necessary transmission infrastructure.

Meanwhile, AMD CEO Lisa Su (https://www.amd.com) will stress the importance of building data centers rapidly and powering them with reliable, clean energy. She emphasizes that AI should extend beyond the cloud to ensure that it is integrated into everyday devices, making it as accessible as electricity.

In a bid to accelerate AI development, Brad Smith calls for the government to open its vast databases for AI training. He argues that the U.S. government is a largely untapped source of high-quality and high-volume data, and making this information available could significantly enhance AI capabilities. Both Smith and Intrator agree that without a proactive approach to managing energy and data resources, the U.S. could risk losing its competitive edge in the AI landscape.

Cited: https://www.reuters.com/business/microsoft-urge-senators-speed-permitting-ai-boost-government-data-access-2025-05-07/