On 10^th of December 2025, Satya Nadella, CEO of Microsoft announced about the Microsoft investment of 17.5 billion dollar in India to build the digital infrastructure, this boom is continuous and because of the adoption of AI, cloud computing, and to preserve the data sovereignty which mandates that the data will be kept within the national border. Because of the rapid expansion of the infrastructure, India’s data centers capacity is projected to increase from 1.2 GW in 2025 to 17 GW by 2030. According to OpenAI researchers, since 2012 the amount of computing power required to train cutting-edge AI models has doubled in every 3.4 months. The Government sees this infrastructure such as data centres as a necessary evil and a very strong pillar of the digital economy, for which the government even accorded infrastructure status to companies to facilitate much easier credit and foreign investment.

Although the global tech companies are pledging for a net zero goal, the other side of the story is completely different, in India 78% of power generation is from fossil fuel based, which is largely coal. So, the expansion of data centres will lead India to emit higher carbon. This further exaggerates by the higher use of AI models which consumes around 10 times more electricity per query in comparison to a standard Google search and even for taring a single large AI model can emit roughly five times the lifetime emissions of an average car. But what about the environmental impact of this growing digital infrastructure, are we regulating it or not? The answer is somewhere between yes or not, because, we do not have unified framework to regulate it instead the environmental governance of digital infrastructure is fragmented across multiple regulatory domains because of which each domain addressing a limited aspect of it, and to take measures to reduce carbon footprint largely rests on the voluntary initiatives by the private companies only.

High Water and Energy Use by Data Centres: Absence of Clear Legal Control over Data’s Environmental Footprint

Data centres are water intensive facilities which require water for cooling of their systems to prevent servers from overheating, even a modest 100 MW data centre consumes around 2 million litres of water daily which is equivalent to the needs of about 6,500 households. This thirst increased by the rise of AI, a simple conversation of around 20-50 questions consumes approximately 500 ml of water, and training of large AI models could consume up to 7 lakh litres of fresh water. A projection showing that the India data centers are projected to increase its capacity to 17 GW by 2030, which would require hundreds of billions of litres of water annually. In cities such as Mumbai, Bengaluru, and Chennai, data centre consumption could outstrip the drinking water demand by more than six times.

India lacks the cohesive legal framework specifically target the environment impact of data centres, current laws such as Environment (Protection) Act, 1986 or the Water (Protection and Control of Pollution) Act, 1974, do not address the digital carbon footprint or the hydrological impact of high-density computing facilities, while the Information Technology Act, 2000 governs digital transaction but it does not govern or address environment sustainability. In the absence of mandatory regulation, monitoring of uses of water is largely voluntary and inconsistent, major contributors such as AdaniConneX and Nxtra by Airtel often do not publish standalone ESG reports regarding their data centres make impossible to segregate the impact of data centres along with India do not have any notified metric for water usage effectiveness. In Greater Noida, the Yotta Data Centre Park operates in groundwater stressed notified areas, although the company claims that they use treated water, but government records remain opaque regarding specific water consumption and groundwater permission.

The current legal regime regarding data’s environmental footprint is more or less in a limbo, because general sustainability reporting frameworks do exist but specific laws that mandates the disclosure and mitigation of the energy footprints and carbon footprints of data centres and AI remains underdeveloped and voluntary. The current legal regime treats environmental emissions as byproducts of traditional industrial production such as steel or oil rather than computation. This makes the AI and data centre operations fall into a legal vacuum.  It is also a noticeable point that the Governments across the globe are racing to develop their national AI strategies and rarely take sustainability and environment into account and most countries’ regulations regarding AI and data centers are centered around data privacy, security, and algorithmic bias, and not focusing on the issue of environmental risks. Lack of binding statute, or standardized reporting, allows companies to conceal specific carbon costs of high-intensity AI models which is also known as carbon-laundering.

Fragmented Environment Governance of Digital Infrastructure in India

Although the Energy Conservation Act, 2001 establishes the Bureau of Energy Efficiency which promotes energy efficiency standards and digital infrastructures also falls under the purview of Energy Conservation Building Code, which sets minimum requirements for energy-efficient design and construction. The MeitY released a draft Data Centre Policy in 2020 to establish data centre zones and streamline the regulatory compliance, but this policy remains only on paper. Parallelly, the Indian Green Building Council operates a Green Data Centre Certification Programme that rates facilities from bronze to platinum on the basis of energy efficiency and sustainability criteria, and promotes an ideal power usage effectiveness (PUE) of less than 1.4 and the use of renewable energy. While a new draft Data Centre Policy 2025 is currently being circulated among the stakeholder to propose tax holidays and harmonized regulation, it has not been enacted yet. 

In absence of any central law, the regulatory framework is fragmented across different states. States such as Maharashtra, Karnataka, Tamil Nadu, Telangana, and Uttar Pradesh have formulated their data centre policies with varying incentives and compliance requirements. Environmental clearances are granted by State-Level Environment Impact Assessment Authorities (SEIAA) but monitoring and transparency of this assessment is often weak, as seen in the Yotta Data Centre. This non-uniformity could affect operators as they have to navigate through the complex maze of rules in every jurisdiction.

Ensuring Sustainable Digital Growth: Need for Legal Reform

To ensure sustainable digital growth in India and the rapid expansion of AI and data centres. Legal reforms are much needed, and we need to move beyond voluntary sustainability to a mandatory obligation which is necessary for environmental accountability. Present laws focus on data, security, and transactions, which is only one side of the digital infrastructure but they do not address the other side compromising of environmental harm caused by digital infrastructure. 

Firstly, transparency must be made compulsory as the data centres consume large amounts of electricity and water, yet their actual usage is rarely disclosed. The companies need to report standard indicators such as power and water efficiency. Also the AI systems should be treated separately, with mandatory disclosure of emissions from training and operating of AI models. Independent audits should verify these claims to prevent whitewashing. Secondly, water protection is critical because many data centres operate in water-stressed cities, and to save cities from water outstripping the restriction on new centres in such areas should be put and ongoing the data centres required to use non-potable water only such as treated wastewater. The companies must also enter binding agreements with local communities to protect water access and livelihoods which will be helpful in protecting constitutional ethics. Thirdly, energy regulation must be strengthened, and the use of renewable energy should be made compulsory, not optional, and data centres should be assessed for their impact on electricity grids and incentives should be rewarded to efficient cooling and energy-saving technologies. Regulating digital pollution with outdated laws is like controlling factory pollution using traffic rules. Long story short, digital growth needs modern environmental laws.