The intelligence-energy equation: Why AI and clean energy cannot thrive without each other

The factory floor has always told a story. 

Once, it was the sound of human hands assembling components, learning by repetition. Then came automation, machines repeating tasks with tireless precision. Today, a quieter but far more powerful shift is underway. The factory is beginning to think. 

Artificial Intelligence has moved from boardroom buzzword to operational reality. It is no longer about futuristic promises; it is about decisions being made in milliseconds, inefficiencies spotted before they occur, and systems that learn continuously. Rapid advancements have made AI-enabled processes and applications central to the energy manufacturing domain. AI is becoming the modern factory’s nervous system, determining how each process will interact and affect the outcome of the other.  

What AI really means for the energy industry 

Over the past five years, AI has moved from experimentation to execution across the global energy industry. Its impact is now measurable across all functions of business, such as operations, markets, tendering and infrastructure.  

AI-led quality inspection and process optimisation have transformed manufacturing. AI-enabled grid management, including fault detection, congestion management, and voltage control, have become essential as distributed energy resources scale.  

Across the energy manufacturing chain, AI is amplifying impact end to end, from improving module manufacturing yields and predictive maintenance, to enhancing solar forecasting accuracy, optimising plant performance, and extending asset lifecycles. In an industry defined by scale and cost discipline, intelligence is becoming a competitive differentiator. It’s allowing business enterprises to tap unexplored avenues and unlock innovation while simultaneously reducing the cost and time of implementation. AI-driven demand forecasting has significantly improved grid planning and dispatch. According to the International Energy Agency (IEA), advanced forecasting models using machine learning have reduced forecasting errors by up to 20–30% in high-renewable grids, helping system operators manage variability more effectively.  

Predictive maintenance has also become a critical lever for asset reliability. McKinsey estimates that AI-enabled maintenance can reduce unplanned downtime in production activities by 50% and extend equipment life by 20–40%, directly improving project economics across generation and manufacturing assets. Overall maintenance costs can be reduced by 18-25%, improving the project’s profitability. 

The International Energy Agency estimates that in the absence of digital technology-enabled grids, losses in electricity supply could amount to almost $1.3 trillion through 2030. The said lost revenue could seriously affect utilities' finances and hamper economic development. In some countries, unreliable grids have been known to result in gross domestic product (GDP) losses of up to 6%.  

The responsibility that comes with intelligence 

AI’s rapid adoption in energy and manufacturing brings undeniable benefits, but it also has a material environmental cost that must be acknowledged. The very systems that are making factories smarter and operations more seamless also require vast amounts of energy, much of it still sourced from fossil fuels today, thereby contributing to pollution and climate change.  

Data centres, the backbone of AI computing, are already significant energy consumers. In 2023, these data centers accounted for roughly 1-1.5% of global electricity usage and about 1% of energy-related CO₂ emissions - a footprint comparable to entire industrial sectors, and these figures are growing with expanded AI workloads. The projected emissions from the top 20 AI systems could reach up to 102.6 million tons of CO2 equivalent, surpassing the emissions from some traditional sectors.  

The scale of energy and carbon emissions tied to AI is not trivial. Training and running large models like generative language models consume substantial electricity. Training a single advanced AI model has been estimated to generate emissions comparable to powering hundreds of homes for a year. That is because much of the power used by data centres is still derived from fossil fuels in many regions, meaning that as AI use scales, so too can associated greenhouse gas emissions and air pollutants that are harmful to both climate and health. Study shows that global data centres consumed 415 terawatt-hours (TWh) of electricity in 2024, with a 12% annual growth rate. At this rate, the figure is projected to exceed 1,000 TWh by 2030, equivalent to Japan’s annual electricity consumption in 2025. The need for a clean energy alternative is evident and inevitable.  

Intelligence and Energy: Resolving the AI–Climate Paradox 

This presents a defining paradox of our time. And, the next industrial revolution will not be defined by intelligence alone, but by how cleanly that intelligence is powered. AI is positioned as a powerful accelerator of decarbonisation, yet without a clean energy foundation, its rapid growth can itself drive emissions higher. The solution does not lie in slowing innovation, but in powering it responsibly with clean energy, energy-efficient hardware, and smarter algorithm design so that the environmental cost of intelligence is minimised.  

In this context, clean energy and AI are no longer separate conversations. AI needs clean power to scale sustainably. Clean energy needs intelligence to scale efficiently. Factories being built today are being designed not just for today’s demand, but for tomorrow’s AI-powered energy economy. 

The choices we make today about technology, energy use, and scale will shape how responsibly the global economy grows tomorrow. AI is rewriting the economics of clean energy manufacturing. Our responsibility is to ensure it does so in a way that is sustainable, resilient, and aligned with the future we want to build. In the end, the intelligence powering tomorrow’s factories must itself be powered by clean energy.  

(The author is Chairman & Managing Director, Vikram Solar. Views are personal.)