Economics Of AI Depends On Nuclear Power | The Reason Why

Artificial intelligence is in the news every day for different reasons. Some ask whether it is a bubble, while others worry about their jobs. Answers to these questions aren't clear yet. However, we know one thing for sure, that it has become an important part of our lives, directly or indirectly. As we look beyond the computing power and advanced algorithms, it becomes evident that this competition will trickle down to something basic like reliable electricity.

Need For More Electricity

According to the International Energy Agency (IEA), data centres consumed roughly 415 terawatt hours (TWh) of electricity in 2024, about 1-2% of global demand. IEA and Deloitte project that it could reach roughly 900-1,100 TWh by 2030. Grids in many regions may struggle to meet this surge in demand without generating new capacities. Much of this growth is expected to come from AI, which is far more energy-intensive than traditional cloud computing. Therefore, we need a higher, continuous, stable and reliable supply of electricity. Certain operators are investing in backup generators for continuous electricity, but these rely on fossil fuels.

Why Nuclear Energy Could Be The Future?

Unfortunately, relying entirely on renewables becomes expensive because of intermittency and the need for large-scale storage. They even need extensive land and specific environmental conditions, which aren't always available everywhere. So experts like Anas Alhaji, an energy economist, think we will use more natural gas in the near term to cope with rising energy demand. But over the longer term, we need a more sustainable solution and nuclear power is increasingly viewed as one of the few scalable low-carbon sources.

Nuclear plants can operate at capacity factors of 80-90%, meaning they produce electricity close to their maximum output for most of the year. In comparison, solar operates at about 15-25% and wind at 25-40%, because their generation depends on sunlight and wind conditions. Even nuclear power needs about 30 times less land than solar and over 170 times less than wind to produce the same amount of electricity, making it easier to connect cities and data centre clusters. However, building nuclear reactors is a time-consuming task and costs billions of dollars.

How Are Countries Securing Nuclear Supply Chain?

Technology companies are already moving to secure nuclear electricity. Microsoft, Google, Amazon, and Meta have signed agreements with Vistra, TerraPower, and Oklo to supply power for future AI data centres.

At the same time, the global nuclear sector is entering a new phase of strategic competition. Data compiled by the Energy for Growth Hub shows that since 2023, Russia has signed about 28 nuclear cooperation agreements, the US roughly 23, and China around 14. Countries like Canada, South Korea, Japan and India are signing new agreements to secure their supplies and build more reactors.

India's Preparedness

In December 2025, India passed the Sustainable Harnessing and Advancement of Nuclear Energy for Transforming India (SHANTI) Bill, which overhauls the legal framework governing nuclear energy and opens the sector to private and foreign players.

As of 2025, India's installed power capacity is about 514 GW, with nuclear contributing roughly 8.8 GW, or less than 2%. India aims to achieve 100 GW of nuclear capacity by 2047. Considering the time and money required to build nuclear power stations, this is very ambitious.

A few days ago, it signed a uranium supply agreement with Cameco, a Canadian company. In December, Russia agreed to cooperate on multiple projects, including the expansion of the Kudankulam plant.

Apart from this, India holds the world's largest thorium reserves, estimated at 846,000-1,070,000 tonnes in monazite sands along the Kerala coast. But here's something unique about thorium-based nuclear reactors. Not many know how this works.

In the 1960s, Oak Ridge National Laboratory in the US developed thorium-fuelled reactors but later abandoned the project. China has since built a prototype in the Gobi Desert and is investing heavily to scale the technology. Eric Townsend, host of the Macro Voice Podcast, cautions that China's success in mass-producing small breeder reactors could give it long-term dominance in global energy, while the West will fall behind.

That means, if India uses thorium, it could reduce dependence on imported uranium. Plus, it could gain an edge by exporting material and technology to others. However, such a transition remains decades away. For that, India must start laying the foundations today.

Final Take

AI is now more about energy, and thus securing energy is a must in the AI economy. For India, this involves boosting nuclear capacity and exploring options like thorium energy.

Even if AI does not scale as rapidly as expected or becomes more energy efficient over time, a secure nuclear power supply will remain valuable. It can help countries reduce dependence on fossil fuels, strengthen energy security, and meet climate targets. Having said that, we need to be extra cautious in dealing with political and public consensus, waste management and reactor safety.

Disclaimer: The views expressed in this article are solely those of the author and do not necessarily reflect the opinion of NDTV Profit or its affiliates. Readers are advised to conduct their own research or consult a qualified professional before making any investment or business decisions. NDTV Profit does not guarantee the accuracy, completeness, or reliability of the information presented in this article.

ALSO READ: AI-Native Banking: Why India's Advantage Is Structural, Not Incremental