As artificial intelligence (AI) and its attendant data demand continue to expand in India and worldwide, a curious dilemma has arisen: will AI help transform energy delivery for the better or will the data centres crucial to its operations impose a new burden on the world’s power grid?
In a 2024 report, the International Energy Agency (IEA) highlighted the growing interconnections between energy and AI worldwideIt projected that demand from data centres would more than double by 2030 to around 945 TWh and that AI would be the principal driver. The demand from AI-optimised data centres was projected to more than quadruple by 2030.
A McKinsey report has also estimated that the annual global demand for data centre capacity could rise at 19-22% from 2023 to 2030, reaching 171-219 GW, against the total current demand of 60 GW. To avoid a deficit, at least twice as much data centre capacity built since 2000 will have to be in place in less than a quarter of the time.
Given AI’s significant hunger for computing power, energy demand is naturally increasing, Ms. Anwesha Sen, an assistant programme manager at Takshashila University studying the impact of technology policy and AI on society, said. She is however optimistic that it’s “not as drastic when compared to other energy-intensive industries”.
Worldwide, data centres consume 1-2% of total power and that’s expected to increase to 3-4% by 2030. To compare, the steel industry consumes around 7% of total power, Sen said.
Pressure, and potential
According to McKinsey, India’s data centre demand is projected to increase from 1.2 GW in 2024 to 4.5 GW by 2030, driven largely by AI and digital adoption across sectors.
Mumbai accounts for 41% of the data centre capacity, followed by Chennai (23%) and the National Capital Region (14%).
AI-driven data centres in India are projected to consume an additional 40-50 TWh of electricity annually by 2030, according to Raghu Raman, professor and Dean at the School of Business at Amrita Vishwa Vidyapeetham.
The increasing adoption of AI and digital technologies in India is contributing to a significant rise in energy demand, especially in already energy-intensive sectors like real estate, Vimal Nadar, national director of research at the Mumbai-based India office of Colliers, a global investment company, said. India is the third-largest energy consumer worldwide, after China and the US, with coal, crude oil, and natural gas comprising the bulk of its energy mix.
The energy consumption of data centres is imposing huge pressure on energy systems worldwide, Anish De, global head for Energy, Natural Resources, and Chemicals at KPMG, said, adding: “India will not be any different.”
According to Sen, an equal concern is the correspondingly increasing demand for freshwater required to cool the servers in these data centres.
That said, there is scope to press AI to the service of smarter energy management as well.
“AI is playing a pivotal role in transforming how energy is delivered, utilised, and managed, both globally and within India,” Mr. Nadar said.
On the one hand, AI could help develop energy transition technologies and as well as new materials that mitigate India’s dependence on critical minerals it currently has to import from abroad, Dr. De said by way of example.
“It will also aid faster project development. This is already playing out in the main geographies and will propagate to others quickly,” he added. “We will see energy efficiency and resource efficiency gains that will also be substantial, though not enough to offset the demand. AI itself will support the gains in expansion of clean energy.”
On the flip side, carbon emissions will also increase. “Despite best efforts it is practically impossible to meet this demand from renewables, both from quality and quantity standpoints,” according to Dr. De.
The IEA also noted in its report that AI “could intensify some energy security strains” as “cyberattacks on energy utilities have tripled in the past four years and become more sophisticated because of AI” even as AI tools are becoming critical for energy companies to defend against such attacks.
Renewables rescue
As energy demand intensifies, real estate stakeholders are increasingly prioritising energy efficiency, sustainability, and emission reduction in both new developments and retrofitting of existing assets, Mr. Nadar said.
“Concurrently, there is a growing emphasis on renewable energy adoption. Real estate developers are increasingly incorporating rooftop solar solutions and solar-integrated building systems, further reducing the sector’s reliance on conventional energy sources.”
The IEA has also said a range of energy sources will be tapped to meet data centres’ rising electricity needs although, according to its report, “renewables and natural gas are set to take the lead due to their cost-competitiveness and availability in key markets.”
India and many other countries are taking advantage of AI to enhance energy efficiency and promote sustainable real estate practices, per Mr. Nadar. In India, the Energy Conservation Building Code and the Roadmap of Sustainable and Holistic Approach to National Energy Efficiency schemeaim to integrate AI and data analytics into smart metering, renewable energy management, and sustainable building design.
Also within the real-estate sector, AI-driven solutions like smart lighting systems, predictive HVAC optimisation, and automated building controls promise to reduce energy consumption by up to 25%. Green certifications such as GRIHA and LEED further encourage AI-based monitoring of energy and resource usage.
Data centres are also adopting AI to optimise cooling systems and server utilisation. As of April 2025, nearly one-fourth of the country’s total data centre capacity in major cities had been green-certified, reflecting an explicit focus on creating sustainable infrastructure. Almost 67% of the Grade A office stock across India’s top seven cities is also green-certified.
‘Need some nudging’
Under the National Smart Grid Mission, AI-enabled systems manage demand and integrate renewables, enhancing grid reliability while reducing wastage, according to Raman. The Nxtra (Airtel) Data Centres uses AI-powered cooling and predictive analytics to cut energy use, paired with renewable power purchase agreements to run green data centres. BrightNight’s PowerAlpha AI deployed in India to forecast and optimise hybrid solar-wind-battery plants and ensure 24/7 access to renewable energy while minimising grid stress.
Tata Power ReNew Power and Hindustan Zinc both use AI for real-time load forecasting, reducing outages and optimising power supply in Mumbai, Dr. Raman added. BESCOM in Karnataka has also started using AI to detect faults and ‘heal’ grid sections and thus mitigate downtime. Similarly, smart meters in Uttar Pradesh have been using AI to detect power theft as well as manage demand-side issues.
“A digital energy grid approach aims to build a unified and interoperable power infrastructure, and its potential can be amplified using AI,” Ms. Sen said.
She added that companies are also working to develop “sustainable AI” that uses recycled water and has higher power use efficiency.
“As the race to build the most capable AI systems has got companies investing in massive data centres, a transition of the energy grid itself to use more sustainable power sources is required and might need some nudging by governments,” Ms. Sen said.
T.V. Padma is a science journalist in New Delhi.