How is the AI boom changing the energy world?

Artificial Intelligence (AI) is driving a massive surge in electricity demand — especially from data centres, but is also emerging as a critical tool for solving energy challenges. While training and using AI models like ChatGPT and Gemini consumes significantly more power than conventional computing, AI is also transforming how we produce, distribute, and consume energy. The blog explores how AI is transforming the energy world. Artificial Intelligence (AI) is quickly changing our world. It's in our phones, helps with medical discoveries, and runs complex financial systems. But this digital progress has a hidden cost: it uses a lot of energy. “A single ChatGPT query consumes nearly 10x more electricity than a standard Google search,” according to Goldman Sachs. This significant demand is also leading to higher reliance on renewable energy sources, such as solar and wind power, thereby creating an opportunity. Beyond just energy consumption, AI is poised to transform (and increase efficiency) how the entire energy industry operates, from production to distribution. AI is responsible for energy demand and its efficient use, hence it becomes important to read why and what about this. Why does AI need so much energy? AI works by doing complex calculations. Training and running advanced AI models, especially large ones, requires huge amounts of computing power. AI needs special computer chips called GPUs and ASICs, which are very good at calculations, but they also use a lot more power than standard computer chips. Nvidia's flagship product, a single rack of GPUs, will consume as much as 120 kilowatt hours per day or the equivalent of 4.5 US homes. GPUs consume 100x more power than normal CPUs. The AI-Led electricity demand According to the International Energy Agency (IEA), data centres, cryptocurrencies, and AI used about 460 terawatt-hours (TWh) globally in 2024 (about 2% of the world's use). The IEA predicts this could double by 2030, exceeding 1,000 TWh (4% of world use), and 1,300 TWh by 2035. AI will be the most significant driver of this increase, with electricity demand from AI-optimised data centres projected to more than quadruple by 2030. The expected consumption is slightly more than Japan's total electricity consumption today. The incremental energy consumption from the data centre is almost 7 years of electricity consumption done by 4 major tech companies (Amazon, Google, Meta and Microsoft) According to the IEA, training large-scale AI models like ChatGPT, Gemini, and Notebook LLM consumed approximately 1,700 TWh of electricity. This accounts for around 0.001% of total global electricity consumption across all sources, and nearly 0.1% of the total electricity consumed by data centres worldwide between 2020 and 2024. AIEnergy1.png The high energy use of AI is a big topic in the industry. As Mr. Amit Paithankar, CEO of Waaree Energies Limited, said in their earnings call on April 23, 2025, "One search on Google Gemini or ChatGPT in terms of watts is 10x more intensive than a plain Google search today…even plain Google Search became 10x more intensive than a year because Gemini is embedded into it" This shows how much more energy AI needs for each task. Size of opportunity While it's hard to estimate overall opportunity size, considering electricity generation and capacity building has a varied way of generation and depends on the region it operates in. Considering the global average wholesale electricity price of $170 mn per 1 TWh as per the International Energy Agency, the opportunity size from just generation could be almost $92 bn annually just from generation. Projected impact of AI on global electricity consumption and market opportunity | Description | Value | |--------------------------------------------------|:---------------:| | AI-led electricity consumption in 2024 | 460 TWh | | Projected AI-led electricity consumption by 2030 | 1000 TWh | | Incremental power demand | 540 TWh | | Global average electricity price | $170 mn per TWh | | Estimated market opportunity | $91,800 mn | Source: International Energy Agency Data Centres: Where AI lives and uses power AI doesn't just exist in the cloud; it runs on physical machines in large buildings called data centres. These buildings are the backbone of the digital world, filled with thousands of computers and storage devices. As AI becomes more common, the need for data centres is growing rapidly, especially for very large ones built by big tech companies. Data centres need a lot of energy for more than just running computers, for cooling, uninterrupted power supply, etc. Who will lead in data centre-led electricity consumption growth? The United States and China combined account for 80% of the growth in data centre consumption. AIEnergy1.png The data centre total installed capacity in India is set to double by 2030. Electricity consumption from data centres is contributing to India’s electricity demand growth at a time when India is already among the world’s fastest-growing electricity markets. AIEnergy1.png Even though China and the US dominate the data centre and its upcoming, India will also see its data centre capacity increase from 2GW to 4.5GW by 2030. How did electricity demand stack up against other sectors? Despite the strong increase, data centre electricity demand growth accounts for less than 10% of global electricity demand growth between 2024 and 2030. Other key drivers, such as industry output growth and electrification, the deployment of electric vehicles and the adoption of air conditioning, lead the way. AIEnergy1.png How AI will transform the energy industry Beyond its direct energy use, AI is set to revolutionise the energy sector itself. While AI's impact on energy consumption is noticeable, its ability to improve how energy is made, traded, and delivered could be even more significant. AI technologies can help forecast future energy needs, make traditional energy sources more efficient and less carbon-intensive, find the best locations for new energy projects, deliver energy exactly when and where it's needed, speed up learning from experience, and discover new materials. S\&P Global introduced a framework to sort these opportunities, which can act as a guide for how the industry might adopt AI. Key opportunities include Energy efficiency AI is very effective at making industrial equipment and processes use less energy. Machine learning helps calculate the best settings for machines like compressors and turbines to run efficiently in different conditions without losing productivity. Companies are achieving 5%-8% energy efficiency improvements this way, which directly means similar reductions in emissions. Grid planning and management The electric power grid is a clear example of the complexity that comes with the energy transition. Efforts are underway to use AI to improve how grids are planned and operated. Using AI to make grid planning simulations faster and more efficient can help consider more scenarios and lead to more reliable results, even with more renewable energy. For example, Chile’s transmission operator is using AI to increase simulation speeds by 86%. New material discovery The energy transition relies partly on new materials—from catalysts that lower hydrogen production costs to new battery chemicals that help diversify supply chains and extend storage. Ex - Georgia Institute of Technology’s work with Meta to find new carbon capture materials. Faster learning curves Industrial development often sees costs drop over time as we "learn by doing." The energy sector has seen this in traditional areas (like cheaper unconventional wells) and new clean energy (like the 25% reduction in battery costs between 2019 and 2024). Companies are now using AI to speed up this learning in new energy areas where there are fewer projects. Leading tech companies are making significant investments in AI infrastructure to meet surging demand and improve energy efficiency. Microsoft plans continued capital expenditure through fiscal 2026 but faces challenges with AI demand outpacing infrastructure expansion. CFO Amy Hood mentioned AI capacity constraints beyond June due to faster-than-expected demand growth. Google is ramping up its investments, with $17.2 billion in Q1 2025 for AI-related infrastructure, and plans to increase capital spending by over 40% to $75 billion in 2025. Nvidia is addressing AI’s energy intensity with its Blackwell GPUs, which are 20x more efficient than traditional CPUs. Its data processing units (DPUs) cut energy consumption by 25%, and the company aims to power all operations with renewable energy by year-end. Amazon Web Services (AWS) is seeing rapid growth in AI revenue, prompting efforts to reduce costs by developing more efficient chips, as CEO Andy Jassy noted that AI can be made more affordable through better hardware. Google's DeepMind has shown that AI can reduce the energy used for cooling its data centres by up to 40%, proving AI can help solve the energy problem. Indian companies NTPC, India's largest power utility, is investing in AI and machine learning to optimise grid management and predict equipment failures, supporting its ambitious renewable energy expansion goals. As of 2023, NTPC has introduced artificial intelligence (AI) and machine learning (ML) technologies in its operations to optimise grid management and predict equipment failures. Tata Power, in collaboration with TCS, co-created EnerUni, an integrated energy management platform, to optimise energy systems using AI and digital twins, aiming to reduce operational costs and emissions. Adani Green Energy's AI tools have boosted solar plant efficiency by 12%. Schneider Electric's AI-based HVAC systems implemented in Indian offices have reduced energy consumption by 30%, saving 1.2 million kWh annually. Power Grid Corporation of India's AI-based cybersecurity system thwarted 1,200 attacks in 2024, protecting critical infrastructure. Anant Raj, a data centre player in India, is scaling its DC capacity from currently at 6MW to 307 MW by 2031. JSW Energy CEO said AI will partly drive growth in Indian energy demand. Waaree Energies sees strong US growth driven by AI-led data centre demand, infrastructure upgrades, and industrial electrification trends. The company recently started a solar module manufacturing facility in the US to capture opportunities emerging from AI and data centre-led energy demand. In closing The intersection of AI and energy presents both monumental challenges and unparalleled opportunities. AI is simultaneously a massive energy consumer, driving unprecedented demand for data centre power, and a powerful enabler of efficiency, grid optimisation, and renewable integration. The future energy landscape will be defined by how effectively these two forces are managed. Disclaimer: This article is for informational purposes only and must not be considered investment advice. Investors should consult with experts before making any investment decisions.