by: Amanda Katili Niode, Ph.D.*
Despite its many benefits, AI also has environmental impacts that cannot be ignored, especially regarding energy consumption.
Artificial intelligence (AI) is increasingly being used to address environmental challenges. This technology allows systems or machines to mimic human intelligence to analyse data on a large scale, recognise patterns, make decisions, and complete tasks automatically. With these capabilities, AI opens up huge opportunities in climate change monitoring, energy transition, and waste management. However, behind the benefits, environmental consequences need to be looked at.
One of AI’s most prominent contributions is monitoring and predicting climate change. By analysing extreme weather patterns, such as storms, heat waves, and floods, AI helps scientists more accurately understand climate trends.
In the energy sector, AI plays a role in managing smart power grids that enable real-time energy distribution. This system helps reduce energy wastage while maximising the utilisation of renewable resources. In addition, AI also supports the clean energy transition by assisting the integration of green energy into the previously fossil fuel-dominated electricity grid.
Apart from energy, AI has also made a significant impact in the agricultural sector. Farmers can optimise water usage, fertilisers, and pesticides using soil and weather data analysis. As a result, agricultural productivity increases without sacrificing the balance of the ecosystem. In the field of waste management, AI can recognise and separate waste types more accurately, improve recycling efficiency, and reduce the amount of waste that ends up in landfills.
All these innovations show that AI offers concrete solutions to environmental challenges. However, its effectiveness depends on how data is managed, how infrastructure is built, and how policies align with sustainability goals.
Despite its benefits, AI also has environmental impacts that cannot be ignored, especially regarding energy consumption. Training large-scale AI models require very high computing power, contributing to carbon emissions that trigger climate change.
For example, training a Generative Pre-trained Transformer (GPT-3) model is estimated to consume about 1,287 MWh of electricity, equivalent to the annual energy consumption of more than 120 households in the United States. This process allows AI to learn from data and models created by researchers, engineers, or data scientists to perform various tasks, such as understanding language, recognising images, or making predictions. Currently, more advanced AI models such as GPT-4.5 and GPT-5 are under development and require more computing power.
Noticeably, AI energy consumption does not only occur during model training. Whenever a user asks an AI-based chatbot a question, the system must run complex data processing in high-capacity data centres. It is estimated that a single question to ChatGPT consumes about 0.3 watt-hours of electricity, an amount comparable to one Google search or the equivalent of charging a cell phone for 10-20 seconds. On a per-interaction basis, the energy consumption is small, but when multiplied by billions of queries worldwide, the impact is significant.
With the widespread use of AI, sustainability challenges are a significant concern. Efforts to develop more energy-efficient models and increase the utilisation of renewable resources in data centre operations are essential steps in reducing environmental impact. Awareness of this challenge has also prompted world leaders to take concrete steps. At the AI Action Summit in Paris on February 10-11, 2025, countries signed a joint statement on inclusive and sustainable AI development.
The document said AI should be developed with an open, transparent, safe, human- and environment-oriented approach. The digital divide is also a significant concern, with a commitment to expanding access to AI technologies for developing countries. In addition, more collaborative AI governance is also emphasised so that a few parties control innovation and can use it for broader interests.
The countries that signed the statement at the AI Action Summit are also committed to promoting ethical AI regulation, protecting human rights, and considering its impact on the environment and society.
*The author is the Director of Climate Reality Indonesia
This article was first published in Indonesian on GBN.top
The banner image was generated by OpenAI’s DALL·E via ChatGPT (2025)
