No.
Or rather—it can’t on its own.
AI doesn’t have the ability to stop the progress we’re forcing. Human beings are the unrelenting force destabilizing our climate.
So maybe the real question is this:
Will we allow AI to accelerate environmental collapse?
The thought settles heavy over me and I grip the steering wheel a little tighter. Next to me in traffic, the city bus spits gray emissions which dissipate by the time the light turns. From the car speakers, a podcaster’s voice floats into focus:
“This new project—known as Stargate—would consume 15 Gigawatts of power. That’s like 15 new Philadelphia-sized cities consuming energy.” NPR’s deluge of facts rain heavy on my ears, and I feel that familiar tightness pull on my abdomen. Climate doom.
“Stargate plans to invest at least $500 billion to build 20 new large data centers across the country…”
Like many Americans, I’ve come to rely heavily on artificial intelligence to help me puzzle out the world around me. This reliance comes at a cost. AI doesn't exist in a vacuum—it depends on an energy-hungry infrastructure. It leans heavily on energy-intensive systems: 99% of all digital data passes through 1.4 million kilometers of deep-sea telecommunication cables. These systems, which power everything from email to streaming to AI models, demand enormous energy, largely sourced from fossil fuels.
AI models rely heavily on water to cool the large data centers. The World Economic Forum reports that “a 1 megawatt (MW) data centre can use up to 25.5 million litres of water annually just for cooling – equivalent to the daily water consumption of approximately 300,000 people.” To address this, innovators are exploring new cooling technology such as cool-water tank immersion and synthetic liquid coolants as alternatives to heavy water use.
In addition to water, the models need electricity. The Harvard Business Review found that “even putting aside the environmental toll of chip manufacturing and supply chains, the training process for a single AI model, such as a large language model, can consume thousands of megawatt hours of electricity and emit hundreds of tons of carbon. This is roughly equivalent to the annual carbon emissions of hundreds of households in America.”
It’s worth noting for anyone who has lamented that AI will take over the world—it already has—and life persists. AI is technology. Just like the automobile, the airplane, the lightbulb. Despite the environmental impacts of driving, flying, and turning on the lights, we keep using these tools.
These tools have consequences for their use. If the private cost of driving, flying, turning on the lights, and using AI is lower than the social costs, we will happily offload those costs onto future generations and speed up climate change in the process. Common policy tools we have to bring that private cost in line with social costs is through a carbon tax or cap and trade policy that puts a price on carbon emissions. In this way we can pay for our technology usage, rather than dine and dash, and leave our grandchildren to pay the bill.
Governments play a pivotal role in shaping a future where AI development is not only innovative, but also sustainable and inclusive. Policy levers already exist to combat environmental destruction:
Carbon taxes and cap-and-trade programs hold producers accountable for their emissions. To learn more about how these policies work, check out Scioto’s recent breakdowns of cap-and-trade and carbon taxes.
Regulatory standards ensure regular and accurate data collection on emissions and energy use.
Data transparency requirements ensure equitable access to the data used to train AI models. This helps address the information asymmetry associated with the 'black box' nature of many large language models.
Procurement rules are another tool to direct government spending in a way that incentivizes or disincentivizes the behaviors of producers within the AI industry.
AI can’t destroy the environment because it would have to get in line—behind us.
This is the moment of the AI revolution, and we still have our foot on the gas, the electricity, and the water. We have to slow our energy and AI demands in order to treat this planet like our grandchildren will live and thrive here, too.