AI carbon emissions have reached alarming levels, with just ChatGPT generating over 260,000 kg of CO2 monthly, raising urgent questions about artificial intelligence’s environmental sustainability.
The staggering scale of AI carbon emissions has come into sharper focus, with recent data revealing that ChatGPT’s environmental impact rivals that of international air travel.
According to a comprehensive study by KnownHost, the popular AI service generates more than 260,930 kilograms of carbon dioxide each month – equivalent to 260 flights between New York and London.
While individual interactions appear modest, with ChatGPT producing just 1.59 grams of CO2 per page view, the platform’s massive user base of 164 million monthly users results in a substantial cumulative environmental impact. The revelation is noteworthy as the tech industry faces mounting pressure to address its growing carbon footprint.
The AI emissions challenge
The environmental cost of AI extends beyond ChatGPT. Other AI platforms, while showing lower total monthly emissions, actually produce more CO2 per interaction. Rytr, for instance, generates 10.1 grammes of CO2 per page view, while Spellbook produces 6.5 grammes – significantly higher than ChatGPT’s per-view impact.
Behind AI’s carbon emissions lies an increasingly energy-hungry data centre infrastructure. The 2024 United States Data Centre Energy Usage Report [PDF] from Berkeley Lab presents a concerning trend: data centre energy consumption more than doubled between 2017 and 2023.
The surge in power use stems from the constant need to maintain optimal operating conditions, including round-the-clock cooling systems and server operations. The International Energy Agency’s Electricity 2024 report paints a more concerning picture for the future.
Advancing AI technologies, alongside cryptocurrency operations, could double energy consumption by 2026, with data centres accounting for one-third of this increase.
Industry response and innovation
The tech sector has begun responding to its environmental challenges. Daniel Pearson, CEO of KnownHost, said: “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.”
Some promising initiatives include:
- Code optimisation to reduce unnecessary processing,
- Advanced cooling methods, including liquid cooling systems,
- Implementation of more efficient airflow designs,
- Increased adoption of renewable energy sources.
Government action and policy response
The Biden administration recognised the growing environmental challenge and issued an executive order addressing the sustainability of AI and data centres. The order explicitly targeted cleaner power transition and acknowledged the fact that one-third of global data centres are in the US.
In a significant recent development, a $500 billion data centre initiative called the Stargate AI project has emerged, led by OpenAI in partnership with Oracle and SoftBank. Newly-elected President Trump has publicly supported the project, describing it as “a resounding declaration of confidence in America’s potential”.
However, the initiative has generated controversy, with Trump advisor Elon Musk engaging in a public dispute with OpenAI CEO Sam Altman over questions of transparency and the project’s feasibility.
Future sustainability challenges
The intersection of AI advancement and environmental sustainability presents a complex challenge for the tech industry. While AI tools like ChatGPT offer certain capabilities, their AI carbon emissions raise serious questions about long-term sustainability. The sector faces the important task of balancing technological progress with environmental responsibility.
The current AI carbon emissions trajectory suggests that without significant intervention, the environmental impact will only increase as the technology becomes widespread.
Combatting that impact will require coordinated efforts from technology companies and policymakers, and pressure from users to implement more sustainable practices and technologies. Success will depend on continued innovation in energy-efficient computing, sustainable data centre design, and the broader adoption of renewable energy sources.





