Discover top fintech news and events!
Subscribe to FinTech Weekly's newsletter
Read by executives at JP Morgan, Coinbase, Blackrock, Klarna and more
The AI Energy Crisis: Can Sustainability Keep Up?
The rapid rise of artificial intelligence (AI) has triggered a new kind of energy crisis. Data centers, the backbone of AI innovation, are consuming vast amounts of power, forcing major tech companies to rethink their energy strategies. Now, the industry is turning toward hydrogen, nuclear, and geothermal energy to ensure future growth remains sustainable.
With companies like Amazon, Google, and Microsoft expanding their data center capacities to support AI-powered services, energy consumption has soared. These facilities, powered largely by high-performance Nvidia graphics processing units (GPUs), require unprecedented levels of electricity. Experts warn that by 2028, data centers could account for up to 12% of total U.S. energy consumption, a steep rise from 4.4% in 2023.
Hydrogen-Powered Data Centers: A Fast-Track Solution?
ECL, a startup founded by Yuval Bachar, a former executive at Meta, Microsoft, and Cisco, is offering one solution: hydrogen-powered data centers. These facilities promise faster deployment compared to traditional grid-connected centers. Hydrogen-based setups can go live in half the time, addressing the urgent need for expanded infrastructure to accommodate the AI boom.
Currently, ECL operates a small-scale 1-megawatt hydrogen-powered data center in Mountain View, California. While hydrogen is primarily sourced from natural gas—a leading energy source in the U.S.—the company’s long-term goal is to transition to green hydrogen, produced through water electrolysis. However, this shift poses economic challenges, as the cost of generating green hydrogen remains high.
Bachar’s startup plans to build a 1-gigawatt data center in Texas within the next four years, with an eye on utilizing hydrogen pipelines. If successful, this project could set a new benchmark for sustainable AI infrastructure.
Nuclear Energy: A Longer-Term Commitment
While hydrogen offers a quicker fix, nuclear energy holds promise for long-term sustainability. Tech giants such as Microsoft have already committed to working with companies developing small modular nuclear reactors. These reactors provide localized, carbon-free power and offer scalability for future needs.
Microsoft has entered agreements to restart operations at a nuclear facility in Pennsylvania, signaling confidence in nuclear energy’s role in reducing carbon emissions. Amazon, Google, and Oracle are also evaluating smaller nuclear installations as part of their sustainability efforts.
However, building nuclear facilities is a lengthy and complex process. Concerns over safety and radioactive waste management continue to delay widespread adoption. Despite these hurdles, tech companies remain optimistic that nuclear power will eventually become a key part of their clean energy strategies.
Geothermal Energy: A New Frontier for AI Power
Another innovative approach comes from geothermal energy. Google has partnered with Fervo Energy, a Nevada-based startup that uses horizontal drilling technology to extract geothermal energy more efficiently. This method has the potential to generate gigawatts of electricity from a single site, making it an attractive option for large-scale AI data centers.
While geothermal energy is a promising solution, the technology faces significant cost barriers. Drilling underground, especially at the scale required for AI infrastructure, is expensive and requires substantial upfront investment.
The Challenge of Scaling Sustainable Solutions
Despite these ambitious efforts, the gap between energy demand and sustainable supply remains wide. Industry analysts predict that by 2028, U.S. data centers could require between 74 and 132 gigawatts of power. With such staggering projections, companies are under pressure to find scalable and eco-friendly energy solutions.
For now, hydrogen and nuclear projects are moving forward, but the transition to fully sustainable energy will take time. Some experts argue that without immediate government support, including subsidies for green hydrogen and nuclear technology, meeting these energy demands will be a significant challenge.
Tech Giants’ Sustainability Goals: A Race Against Time
Major tech companies have set aggressive targets to reduce emissions. Google aims for net-zero emissions by 2030, while Microsoft plans to become carbon negative by the same year. Amazon has set a 2040 deadline for reaching net-zero carbon emissions.
Startups like Last Energy are collaborating with these tech leaders to integrate modular reactors into future data centers, offering flexible on-site power generation. These partnerships reflect the urgency in the tech industry to balance growth with sustainability.
A Future Powered by Innovation—and Responsibility
The energy needs of AI are pushing the boundaries of existing infrastructure, forcing tech companies to invest in cleaner, more efficient solutions. Hydrogen and nuclear energy, alongside geothermal and solar projects, represent the next steps toward ensuring AI’s future doesn’t come at the expense of the planet.
The race to build sustainable AI data centers has already begun. As companies pour resources into cleaner energy, the world will watch closely to see if innovation can outpace the growing energy crisis—and if sustainability can remain a priority in the age of AI expansion.
