AI’s unregulated growth is single-handedly derailing global climate targets, with thousands of new data centers absolutely devouring energy grids, overwhelming cardon reduction plans, and increasing carbon AI emissions.
Electricity demand – powered by fossil fuels – is creating an insatiable new class of energy consumerism that’s actively pushing global climate goals out of reach. The world is now faced with the choice between progress and the planet’s survival, and ours – by extension.
AI climate control has become directly linked to data centers multiplying to support large language models and cloud computing with experts warning that the electricity surge and water consumption could derail global climate targets.
“Accelerating from deployment to a deeply decarbonized, resilient energy system is proving far more complex than simply adding megawatts,” said vice president for scenarios and technologies at Wood Mackenzie, Prakash Sharm.
Analysts say tracking data center carbon emissions has never been more necessary to understand the real-world impacts.
Emissions for the Largest AI Data Centers
Across the US and beyond, power-hungry data centers are increasing in masses to meet Big Tech’s appetite for infrastructure, increasing the carbon AI footprint. Often powered by fossil fuels, these facilities are reversing years of progress in renewable energy adoption.
According to Wood Mackenzie, almost no major economy including the US, Canada, and Germany is on track to meet its 2030 emissions goals. To curb data center CO2 emissions, states are reassessing sitting and energy mixes.
Center for Biological Diversity reports warn of fossil-powered AI data centers’ projected expansion. Every other sector of the US’ economy would need to cut emissions by 60% to stay within national targets. Advocates argue that streamlined AI data center energy geothermal permitting could reduce reliance on gas.
The International Energy Agency estimates that by 2030, global data centers could consume as much electricity as entire nations like Sweden or Argentina, pressuring grids to efficiently sustain AI.
Training AI models like GPT-4 requires astronomical power—roughly enough energy to keep San Francisco running for three days—and each user query adds to that load. Companies say they are accelerating procurement strategies to ensure more sustainable AI.
“Data center energy demand is threatening to sabotage the country’s already faltering climate goals,” Fleming and colleague Jean Su stated.
AI for Sustainable Development
Still, experts caution against framing AI solely as a climate villain. When harnessed responsibly, it can become a powerful ally in sustainability. AI systems already help optimize renewable energy grids, forecast extreme weather, and monitor carbon emissions in real time, sharpening accountability for AI carbon emissions.
Technology giants like Google and Microsoft have pledged to run on carbon-free energy by 2030, and researchers are racing to develop “green AI” algorithms that use fewer computational resources—efforts that could shrink the carbon footprint of AI.
Some companies are adopting “carbon-aware computing,” scheduling AI training when renewable energy is most available, while publishing clearer lifecycle data on carbon footprint AI.
The challenge, experts agree, lies in balance. The carbon AI potential to fight climate change depends on whether its own emissions can be curbed. As Beck put it, “States are on the front lines right now.”
The future of AI—and the planet—will depend on how wisely we choose to power it, including tighter disclosure of carbon emissions of AI.
Inside Telecom provides you with an extensive list of content covering all aspects of the tech industry. Keep an eye on our Tech sections to stay informed and up-to-date with our daily articles.