Panelists partake in a discussion titled “The AI Race And How To Win It" at the July 15 summit. (Credit: Carnegie Mellon University)
Trump, McCormick showcase AI at summit
Many see artificial intelligence (AI) as the one-stop shop to solving the future’s gloomiest problems; now, Pennsylvania may be at the forefront of the nation’s AI revolution.
In a July summit, Senator Dave McCormick (R-PA) and President Donald Trump tied Pennsylvania’s AI development to fossil fuels by announcing more than $90 billion in private-sector commitments to advance AI infrastructure, data centers, power generation, and workforce training in the state.
The investments’ record-breaking scale positions Pennsylvania as a major player in the AI competition, while presenting concerns about future environmental and infrastructure challenges stemming from a partnership with nonrenewable energy sources.
The inaugural Pennsylvania Energy and Innovation Summit took place at Carnegie Mellon University on July 15. The Republican duo was joined by top executives from the technology and energy industries — including Meta, Microsoft, Google, ExxonMobil, Blackstone, and Amazon — as well as other government officials, such as US Secretary of the Treasury Scott Bessent and Governor Josh Shapiro.
“I am proud to partner with President Trump and the business leaders here today to drive a new era of industrial growth that helps make America energy dominant while creating jobs and opportunities for working families across Pennsylvania,” said Senator McCormick in a release.
From Blackstone’s $25 billion investment for data centers and natural gas-fired plants to AWS’s $20 billion commitment to cloud and AI infrastructure expansion, the investments broke Pennsylvania records as the largest private commitment in the state’s history. The combined investments are anticipated to create tens of thousands of jobs to the Commonwealth.
Some of the projects had already been in the works ahead of the summit, but others were brand new. The overarching theme centered on Pennsylvania as the site of the “next industrial revolution,” emphasizing fossil fuel energy in order to make the country a front-runner in the AI arms race.
One cannot underestimate the energy demands of AI. Latest reports reveal that 4.4% of all energy in the US goes toward AI data centers. Yet, AI is still in its toddler years, and as it progresses, scholars predict that by 2028, AI could consume as much electricity annually as 22% of all US households.
AI and its energy sources are like an inseparable pair. But, to the chagrin of environmentalists, AI data centers rely on constant power, making it harder to rely on volatile — but more sustainable — technologies like wind and solar power. While possible, scholars say it takes more maneuvering. As a result, companies often prefer “dirtier” power sources like coal, nuclear, or natural gas for their data centers.
It is fitting for Pennsylvania to be the site of these investments, given its immense energy production. The Commonwealth is the nation's second-largest natural gas producer after Texas, and the third-largest coal-producing state.
Further, Pennsylvania is a significant producer of nuclear power, which is considered a cleaner energy alternative and is a target for tech leaders. At the summit, executives discussed Microsoft’s collaboration with Constellation Energy to potentially reopen the Three Mile Island nuclear plant.
“As the nation’s second largest energy producer and a global nuclear power leader, Pennsylvania is uniquely positioned to deliver the abundant, affordable energy that growing AI and advanced manufacturing sectors demand,” said Senator McCormick. “We have the skilled workforce to build and operate this critical infrastructure, world-class universities driving innovation, and strategic proximity to over half the country’s population.”
The marriage of fossil fuels and AI at the Carnegie Mellon summit is a logical pairing, but one that is concerning for climate advocates and those with energy expertise.
Panayiotis Moutis is an electrical engineering professor at City College of New York, chair of IEEE-USA’s Energy Policy Committee, and Power & Energy lead at Climate Change AI. In an interview, he discussed the impacts of new nonrenewable energy facilities as they strive to feed power-hungry AI data centers.
Moutis discussed a recent Washington Post report on how AI’s energy demands are causing electricity bills to climb in the Mid-Atlantic as data centers are straining the electrical grid. Despite pledged grid updates from First Energy and PPL Corporation, Moutis said the problem will worsen as energy demand outpaces infrastructure growth.
“What’s happening with the proliferation of gas-fired units is that we are not only committing more fossil fuel energy into operation and into use, but we have also changed our overall electrical grid operating strategy,” said Moutis. “We are at a point that we might not be certain how the grid can maybe survive brownouts, or under similar conditions, even blackouts.”
In regard to the sustainability component, Moutis — who prefers to see a diverse energy portfolio for both environmental and economic reasons — was not shocked to see such a large push for fossil fuels.
He said the Trump Administration has made clear its support for the fossil fuel industry and has followed through, particularly with the One Big Beautiful Bill, which repealed tax credits for renewable energy. “We were not surprised by the Administration, but we were surprised by the big tech companies,” said Moutis, referencing net-zero carbon emissions pledges made years ago by corporations like Meta.
The Pennsylvania investments are part of a broader mobilization for AI development. A week after the summit, the White House released “America’s AI Action Plan,” a 28-page document that features policies enabling innovation and adoption of machine learning.
The Trump Administration has prioritized AI advancement as the solution to progressing past hostile foreign nations, like China. “To remain the leading economic and military power, the United States must win the AI race,” said David Sacks, who holds the newly created AI and Crypto Czar position.
Yet, the full potential of AI has not been realized. Moutis said the AI tools that the electricity industry uses — like load forecasting with neural networks or pattern recognition for satellite images — have been the same for the past 15-20 years.
“So, this is the part where the AI promises are still kind of unfulfilled, but not unrealistic,” Moutis said. “We are holding our breath.”
