By Noleen Mariappen
They Drive You, But They Don’t Drive Themselves
The steering wheels may be turning themselves, but make no mistake: there is a driver, and that driver is artificial intelligence. Every autonomous vehicle (AV) depends on AI to make sense of what’s happening on the road, spotting pedestrians, reading traffic lights, calculating the safest next move. A lot of that brainwork doesn’t even happen inside the car. It happens in giant data centers, scattered around the world, that never stop running. Without those hidden servers, what was the “self-driving” sci-fi future, and now the present, simply wouldn’t exist.
The Illusion of Clean Technology
On the surface, AVs look like the perfect clean-tech solution. No petrol fumes, no tailpipe emissions, no choking city air. But scratch that surface and the story gets murkier. Instead of burning petrol, AVs burn electricity, huge amounts of it, inside server farms that power their AI. We’ve seen this kind of story before. Fossil fuels once felt like unlimited progress, only for the environmental bill to arrive decades later. AVs could be making the same mistake: shifting the pollution out of sight, into racks of servers instead of exhaust pipes.
A Growing Global Appetite for Power
This isn’t just a problem in one region, it’s happening everywhere.
– In the United States, data center construction has exploded. Virginia, the country’s “data center capital,” has already seen power grids stretched to their limits. In Arizona and California, the debate isn’t just about electricity—it’s also about scarce water supplies keeping those facilities cool.
– In Europe, Ireland has become a cautionary tale. Data centers now eat up more than 21% of the country’s electricity, and Dublin has paused new approvals. Across the EU, total demand could nearly triple by 2030, reaching 150 terawatt hours, roughly equal to the entire consumption of Poland.
– In Asia, China is going all-in. Cities like Shenzhen and Beijing are building massive hubs to support national AV and AI ambitions. But since much of the electricity still comes from coal, the “green” credentials are shaky at best. India, meanwhile, is debating how to expand testing of AVs without blowing past its climate targets.
Globally, the International Energy Agency warns that by 2030, data centers could swallow up nearly 1,000 terawatt hours a year, about as much power as Japan uses in total.
The Hidden Thirst: Water
It’s not just electricity. Data centers are also thirsty, very thirsty. Cooling those endless racks of servers takes around two liters of clean water for every kilowatt hour of energy consumed. Training GPT-3, one of the AI models initially underpinning AV technology, used more than 700,000 liters on its own. In water-stressed places like California, northern China, and southern Spain, people are starting to ask whether limited freshwater should be keeping AI servers cool while households are told to conserve.
Pushback and Policy Moves
Governments and communities are beginning to push back.
– The European Union is tying its Green Deal to data transparency, forcing operators to disclose how much energy and water they consume.
– In the United States, federal regulators are investigating how AI-driven demand could destabilize national grids, while local lawsuits in Oregon and Virginia are trying to block new builds.
– China and India are experimenting with efficiency standards, though their rapid expansion makes it hard to keep pace.
At the same time, much of the public debate is still focused elsewhere, mainly on jobs. Taxi drivers, truckers, and delivery workers worry about automation. But there’s another question hiding behind those headlines: what will the environmental costs of AVs be once the servers behind them dominate local grids and water supplies?
A Familiar Story
It all feels eerily familiar. The car once symbolized freedom, until the world woke up to the smog, oil dependence, and climate chaos it left behind. Now AVs are being pitched as clean, modern, unstoppable. Yet the more you look, the more it seems like history is repeating itself: convenience today, consequences tomorrow.
Searching for Solutions
There are fixes on the table. Tech giants are pouring money into renewable energy projects to offset their servers. Startups are testing new cooling methods that use less water, and some governments are talking about “green data center” certifications. But right now, those efforts are dwarfed by demand. AI and AV growth is accelerating far faster than sustainability measures are being rolled out.
The Road Ahead
Autonomous vehicles may one day reduce our reliance on fossil fuels, but they’ve already locked us into a new form of dependency: energy-hungry, water-thirsty infrastructure. This isn’t a European problem, or an American one, or an Asian one, it’s all of them, all at once, and unless we get serious about the costs of the AI that powers them, the “clean” future of transport could end up looking a lot dirtier than we think.
References
MIT News (2023). The carbon emissions of autonomous vehicles. https://news.mit.edu/2023/autonomous-vehicles-carbon-emissions-0113
International Energy Agency (IEA) (2024). Energy and AI: Executive Summary. https://www.iea.org/reports/energy-and-ai/executive-summary
World Economic Forum (2024). Europe data centre demand set to soar. https://www.weforum.org/stories/2024/11/europe-data-centre-plus-other-technology-news-to-know
IFRI (2024). AI, Data Centers, and Energy Demand in Europe. https://www.ifri.org/en/papers/ai-data-centers-and-energy-demand-reassessing-and-exploring-trends-0
AP News (2025). UK faces local pushback over massive AI data centers. https://apnews.com/article/fdb196e2dec8bdf18eab6b8a6a672cbd
The Guardian (2025). Labour’s AI growth zone sparks water shortage fears. https://www.theguardian.com/technology/2025/jan/13/labour-ai-datacentre-growth-zone-water-shortages-abingdon-reservoir
Food and Water Watch (2025). Artificial intelligence, water, and climate. https://www.foodandwaterwatch.org/2025/04/09/artificial-intelligence-water-climate
Yale E360 (2024). The environmental cost of AI. https://e360.yale.edu/features/artificial-intelligence-climate-energy-emissions
_______________________________________
Noleen Mariappen is a purpose-driven impact strategist and tech-for-good advocate bridging innovation and equity across global communities. With a background in social and environmental impact and a passion for digital inclusion, Noleen leads transformative initiatives that leverage emerging technologies to tackle systemic inequality and empower underserved populations. Noleen is an active contributor to ethical AI dialogues and cross-sector collaborations focused on sustainability, education, and inclusive innovation. Connect with her on LinkedIn: https://www.linkedin.com/in/noleenm/
__
The views expressed in this article are those of the author and may not reflect the official stance of Consumer AI Protection Advocates (CAIPA).
CAIPA’s mission is to empower consumers by advocating for responsible AI practices that safeguard consumer rights and interests across various sectors, including electric vehicles (EVs), autonomous vehicles (AVs), and robotics.
#CAIPA #ArtificialIntelligence #ConsumerProtection #AutonomousVehicles #FutureofWork
