ChatGPT's carbon footprint is astronomical and not sustainable
Anyone who thinks booking a commercial flight is the only way to rack up carbon emissions should take a closer look at the monthly output of popular AI chatbots like ChatGPT.
Recent data indicate that the emissions associated with these tools could raise eyebrows among those who care about shrinking our collective environmental footprint.
High carbon footprint of ChatGPT
According to a study by KnownHost, the use of ChatGPT produces more than 260,930 kilograms (573,000 pounds of carbon dioxide (CO2) each month. That figure is linked to high traffic from its more than 164 million monthly users.
Although it emits only 1.59 grams of CO2 per page view, that efficiency is offset by soaring popularity.
The study also points out that ChatGPT’s carbon impact is equivalent to 260 flights from New York City to London in the same timeframe, offering a stark reminder that digital services leave a mark on the environment.
ChatGPT isn’t the only carbon abuser
Other AI platforms show less total monthly CO2, but they emit more per individual pageview. Rytr’s output stands at 10.1 grams (0,35 ounces) CO2 per pageview, while Spellbook comes in at 6.5 grams (0.23 ounces) per pageview.
Even though their overall usage is smaller, the carbon cost per interaction is higher than ChatGPT’s. Some might see ChatGPT’s lower rate per pageview as a bright spot, but its heavy traffic means emissions still climb significantly.
Data centers consume massive energy
Data centers host these AI tools and, according to the 2024 United States Data Center Energy Usage Report from Berkeley Lab, they consumed more than double the energy in 2023 than they did in 2017.
“Electricity consumption of US data centers is currently growing at an accelerating rate,” stated the authors of the report. These massive facilities keep servers cool and operational around the clock, leading to constant energy draw.
The International Energy Agency’s Electricity 2024 report notes that newer technologies – like advanced AI and cryptocurrency – could double energy use by next year. One-third of that rise is expected to be traced back to data centers.
This trend reveals how widespread digital services can become major contributors to global energy consumption.
“As innovation continues into 2025, balancing environmental sustainability will need to become a core focus for tech companies and the users of their tools alike,” said Daniel Pearson, CEO of KnownHost. The statement suggests a need for shared responsibility.
Developers can shift toward greener solutions, and users may opt for options that cut power use.
Governments push for greener AI
The Biden administration has responded with an executive order that addresses the future of AI and data centers. It urges the transition to cleaner power, given that a third of the world’s data centers are in the United States.
The announcement coincides with news of the Stargate AI project, a $500 billion data center initiative led by OpenAI in partnership with Oracle and SoftBank, which President Trump described as “a resounding declaration of confidence in America’s potential.”
However, the project has sparked a feud between Elon Musk, a Trump advisor, and OpenAI CEO Sam Altman over the transparency and feasibility of the investment.
Cutting AI carbon emissions
Some AI tools already push down their energy load by optimizing code and trimming unnecessary processes. Meanwhile, data centers explore better cooling methods, such as liquid cooling and more efficient airflow designs, to cut overall energy consumption.
Renewables like solar or wind power also help lower fossil fuel reliance. Tech companies that concentrate on reducing emissions in hardware and software may find themselves at a competitive advantage, especially as the public eye shifts to climate impact.
AI applications are here to stay, and so is the energy they consume. While ChatGPT’s popularity puts it in the spotlight, other AI tools also contribute to rising energy demands.
Striking a balance between progress and sustainability will require a combined effort from developers, policymakers, and everyday users. Improving efficiency not only benefits the planet but also ensures that these digital services stay reliable.
The study was published by KnownHost.
—–
Like what you read? Subscribe to our newsletter for engaging articles, exclusive content, and the latest updates.
Check us out on EarthSnap, a free app brought to you by Eric Ralls and Earth.com.
—–
