New Delhi, Nov. 8 -- In Finland, the future of artificial intelligence has run into an inconvenient obstacle: a forest, much of it already cleared before mandatory environmental impact assessments were completed. Authorities have ordered a halt to preparatory work at two planned data-centre sites in Muhos and Kajaani until the assessments are completed.
The order, issued to Google's Finnish subsidiary, Tuike Finland, follows extensive forest clearance and concerns over incomplete environmental impact assessments. Reports claim around 330 hectares were cleared or prepared at Muhos, with nearly 200 hectares affected at Kajaani. The company has been directed to suspend activities that significantly alter the environment. The matter has also prompted a criminal investigation into suspected environmental-law violations at the Muhos site.
The intervention has triggered an angry online debate. A critic with the handle barreda_98 argued that the damage had already been done: "They've already destroyed the damn forest. What does it matter anymore?" Another post, from kardioclan, questioned the pride some Indians take in the Indian ancestry of Google's chief executive: "India is proud just because his ancestors were Indians." The swipe at ancestry aside, the issue is much larger than any one person or company.
Beneath the obvious anger lies a question that extends well beyond Finland: when the pursuit of technological progress destroys the natural resources on which communities depend, who is held accountable? The question is particularly relevant to India, where the construction of massive AI data centres is being celebrated as an investment bonanza.
West's Power Crunch
The United States is discovering that artificial intelligence requires more than brilliant engineers, sophisticated chips and ambitious investors. It requires electricity, water, land and an enormous supporting infrastructure. The country's ageing power grid is struggling to accommodate the explosive growth in power demand from data centres. Utilities are facing severe shortages of transformers, switchgear and other equipment. Lead times for high-voltage transformers have stretched beyond three years, with some of the equipment taking more than 160 weeks to arrive.
The scale of the challenge is staggering. Developers are racing to build AI facilities while utilities scramble to expand generation and transmission capacity. Yet projects announced with extraordinary confidence are encountering practical limits: insufficient power, delayed grid connections, rising costs and opposition from communities concerned about water consumption and electricity bills.
Not every proposed data centre will become reality. Some projects may be delayed indefinitely or abandoned as financing, energy and infrastructure constraints collide. The investment frenzy is running far ahead of the physical capacity required to sustain it.
Local resistance is also growing. Communities in several American states have challenged projects over their anticipated demands on electricity and water, as well as their effects on neighbourhoods and local environments. The promise of jobs and investment is regularly being weighed against the costs residents may have to bear.
It is against this backdrop that India's sudden emergence as a preferred international destination for AI infrastructure must be understood.
India: The New Frontier
Global technology giants building the AI economy-Microsoft, Amazon, Google and Meta, among others-need somewhere to expand. They are looking for land, reliable power, water, connectivity, skilled workers and governments willing to facilitate large investments. India offers many of these advantages. Its growing digital economy, engineering talent and large domestic market make it an attractive destination. State governments, meanwhile, are competing to secure projects that promise investment, jobs and a place in the global technology ecosystem.
Andhra Pradesh provides a stark example. The state has allocated around 600 acres across three locations for a proposed Google-linked, 1-GW data-centre hub near Visakhapatnam. The project involves an investment of $15 billion, or Rs 1.40 lakh crore. The numbers may be impressive, but the scale demands scrutiny. A 1-GW data-centre hub is not simply another industrial estate; it represents a vast and continuous demand for electricity, cooling, water, land and supporting infrastructure.
The question is not whether India should welcome investment or participate in the AI revolution; it is whether the country is calculating the full cost of that participation. A project may generate substantial economic activity while placing significant pressure on local resources. Its benefits may be national or global, while its environmental costs are concentrated in the communities surrounding the facility. That imbalance deserves attention before the first server begins humming.
Hidden Green Costs
The environmental footprint of data centres extends well beyond their electricity consumption and carbon emissions. These facilities require cooling systems to prevent powerful computing equipment from overheating. Depending on the technology employed, cooling can consume enormous quantities of freshwater. In water-stressed regions, that demand may compete with households, agriculture and other essential uses.
Land is another critical resource. Large facilities require space for buildings, substations, transmission infrastructure, access roads and backup systems. Forests, agricultural land and natural habitats can become casualties when development proceeds without adequate safeguards.
Research is also beginning to examine the local heat generated by data centres. A 2026 preprint estimated that land-surface temperatures around data centres rose by an average of 2degC after operations began, potentially exposing more than 340 million people worldwide to the effects. The findings point to a possible 'data heat island' effect, although the research is still developing and should not be mistaken for a universal temperature increase across entire regions.
For India, where extreme heat, water scarcity and unreliable electricity remain pressing concerns, these findings cannot be dismissed as distant environmental issues. The challenge is not to reject AI, of course, but to ensure that its infrastructure does not undermine the very communities and ecosystems on which sustainable development depends.
A Different Way Ahead
There are alternatives worth examining. China has launched an underwater data-centre project in Shanghai's Lin-gang area, connected directly to offshore wind power and using seawater for cooling. The technology is not a universal solution. Underwater infrastructure has its own engineering, ecological and economic challenges. But it demonstrates that data-centre design need not be confined to the traditional model of vast buildings drawing heavily on local power and water supplies.
The project's developers say the system reduces electricity consumption by 22.8 per cent, eliminates freshwater use for cooling and cuts land use by more than 90 per cent compared with conventional facilities.
India could explore similar innovations while strengthening environmental assessment, requiring transparent reporting of energy and water consumption, prioritising renewable power and encouraging efficient cooling systems. Data-centre proposals should be evaluated not only by the size of the investment they attract, but by their long-term impact on local resources and communities.
The AI revolution may transform economies and societies. But technological progress cannot be measured only in computing power, investment announcements or the number of facilities built. If the price of AI is forests cleared before assessments are completed, water resources strained and electricity diverted from essential needs, then the celebrated economic gains demand harder scrutiny.
India has an opportunity to become a major player in the AI economy. It also has an opportunity to avoid inheriting the environmental mistakes of others. The real test of progress is not how much infrastructure a country can attract; it is whether that infrastructure leaves the country better off.
Published by HT Digital Content Services with permission from Millennium Post.