A new analysis by environmental researcher Alex de Vries-Gao is raising the alarm on artificial intelligence’s rapidly growing appetite for electricity—and the climate consequences that come with it.
Published in the journal Joule, de Vries-Gao’s study explores how large-scale AI systems like ChatGPT are helping to push global data center energy use to new highs. His estimates show that AI-related computing could consume up to 82 terawatt-hours of electricity in 2025—roughly equivalent to Switzerland’s entire annual power use.
The AI Boom, Powered by Data Centers
De Vries-Gao, a Ph.D. researcher at the VU Amsterdam Institute for Environmental Studies, built his projections using a simple but revealing approach: calculating the energy required to process user queries across various AI platforms.
His work comes as the International Energy Agency (IEA) reports that data centers were responsible for up to 1.5% of global electricity use in 2024. And with demand still climbing, the AI boom is proving to be a key driver—alongside other power-hungry technologies like cloud services and cryptocurrency mining.
Transparency Trouble and Tech’s Energy Footprint
One major concern, according to de Vries-Gao, is the lack of transparency. As AI workloads skyrocket, tech companies have become increasingly tight-lipped about their actual energy consumption—prompting independent researchers to fill in the gaps.
What they’re finding is worrying: if AI demand doubles by the end of the year, it could account for nearly 50% of all energy used by data centers worldwide.
That kind of growth isn’t just a utility bill problem—it’s a climate problem.
From Grid Power to Greenhouse Gases
“AI providers are still relying heavily on grid electricity,” de Vries-Gao notes. And while some companies are experimenting with in-house energy generation, many still tap into power grids dominated by fossil fuel sources.
The result? A surge in greenhouse gas emissions, which could accelerate climate change and potentially drive higher energy prices across global markets.
What’s Next?
While the AI revolution promises smarter tools and faster services, de Vries-Gao’s research adds a critical reminder: progress comes with a power bill. As AI continues to scale, the pressure will mount for companies to find greener, more sustainable ways to fuel the future of intelligence.
And for policymakers and climate advocates, the message is clear—now is the time to ensure AI growth doesn’t come at the planet’s expense.
