Meta Secures Nuclear Power to Meet AI Energy Demands
Meta has taken a significant step to address the surging energy demands of artificial intelligence by entering into a 20-year power purchase agreement (PPA) with Constellation Energy. This deal will supply zero-carbon electricity from the Clinton Clean Energy Center in Illinois, beginning in June 2027, coinciding with the expiration of Illinois’ zero-emission credit program. The agreement includes an additional 30 megawatts (MW) of output, sufficient to power a city of approximately 30,000 people, and is expected to generate an estimated $13.5 million in annual tax revenue while preserving 1,100 jobs in the region .
The Clinton plant, with a capacity of 1.1 gigawatts (GW), will play a critical role in supporting the clean energy needs of Meta's AI infrastructure. The company’s reliance on AI workloads is skyrocketing, leading to unprecedented power requirements something intermittent renewable energy sources alone cannot consistently satisfy. By incorporating nuclear into its energy mix, Meta is ensuring steady, around-the-clock baseload power that aligns with its sustainability goals.
Nick Brown, CEO of Constellation Energy, commented, “This partnership underscores the vital role of nuclear power in decarbonizing the digital economy while ensuring reliable electricity.” Similarly, Meta’s Global Head of Energy, Peter Freed, noted, “We’re committed to sourcing energy that matches the around-the-clock needs of our AI systems. Nuclear power helps us meet this goal while reducing our carbon footprint.”
/*Tech Industry's Shift Towards Nuclear Energy*/
Meta is not alone in this nuclear pivot. Other tech giants such as Microsoft, Amazon, and Google are also exploring or entering into nuclear partnerships. Google has collaborated with Kairos Power to explore SMR applications, while Microsoft has engaged in projects involving the potential restart of a reactor at Three Mile Island . Amazon is working with advanced nuclear developers to power its data centers . These moves highlight a growing industry-wide recognition that AI and clean energy must scale together and that nuclear power may be essential to bridge the gap.
/*Meta's Long-Term Nuclear Strategy*/
In late 2024, Meta issued a request for proposals (RFP) to source between 1 to 4 GW of new nuclear capacity across the United States. The RFP is open to both large conventional reactors and next-generation small modular reactors (SMRs), with deployment targeted in the early 2030s . Meta is not just looking to buy energy it seeks partners capable of managing the full project lifecycle, from permitting and financing to construction and operation. This proactive, long-term strategy signals that the company views nuclear as a core pillar of its future energy architecture.
The RFP welcomes creative cost-sharing arrangements and joint investments, underscoring Meta’s willingness to play a direct role in energy infrastructure development. While the company has already secured over 12 GW of renewable energy contracts to match 100% of its operations, nuclear offers what renewables can’t: reliable and consistent output to support continuous AI workloads and ensure grid stability
/*Challenges and Considerations in Nuclear Expansion*/
Yet, some experts remain cautious. Maria Korsnick, President of the Nuclear Energy Institute, pointed out, “Nuclear power provides unmatched reliability, but the path to new reactors is long and costly. We need policy support and streamlined regulation to accelerate these projects.” Critics argue that while nuclear is reliable, expanding renewable energy coupled with battery storage could be a faster and more cost-effective solution. A 2024 report by the International Energy Agency (IEA) found that utility-scale solar and wind paired with storage can be competitive with new nuclear, especially when accounting for permitting delays and upfront capital costs .
The shift toward nuclear energy comes amid growing concerns about the carbon footprint of tech companies. According to United Nations data, indirect emissions from the tech sector have surged by approximately 150% since 2020, with Meta alone seeing a 145% increase. As AI models become more complex and compute-intensive, data centers are consuming more electricity than ever. While renewables remain an important part of the equation, their intermittent nature makes nuclear a compelling complement especially when stability and scalability are critical.
Still, there are challenges. Building new nuclear plants, especially conventional reactors, involves long timelines, significant regulatory hurdles, and high upfront costs. While small modular reactors promise faster deployment and potentially lower costs, they are still emerging technologies and face their own development barriers. Critics also argue that nuclear is slower and more expensive than expanding renewable energy coupled with battery storage. Nevertheless, Meta appears to be hedging its bets, securing immediate energy needs through gas and solar PPAs while laying the groundwork for nuclear capacity in the next decade.
/*Future Outlook for Meta's Energy Initiatives*/
The Clinton nuclear agreement will officially begin in mid-2027, aligning with the expiration of Illinois’ state subsidy. The longer-term RFP initiative may yield operational nuclear capacity by the early 2030s. In both cases, Meta is proactively investing in infrastructure that not only powers its own needs but also contributes to sustaining America’s clean energy capabilities. These efforts will require policy support, streamlined regulation, and upgraded grid systems to fully realize their potential.
Meta’s nuclear push marks a pivotal moment in how the tech industry approaches energy. With AI’s energy appetite growing rapidly, nuclear power is emerging as a viable, scalable solution to maintain both operational reliability and environmental commitments. By aligning its energy strategy with long-term AI growth, Meta is not just securing power—it’s shaping the future of sustainable computing.