Are AI Generated Images Bad for the Environment? What the Evidence Really Says

AI generated images are becoming increasingly common across entertainment, from character concepts and promotional artwork to virtual performers, story development and independent digital series. They can give creators access to visual tools that once required considerably more time or money, but their rapid growth has also raised questions about electricity, water and the wider environmental impact of generative AI.

The evidence shows that AI images do have a physical footprint, although there is no single environmental cost that applies to every picture. The model being used, image quality, number of generations, data centre and electricity supply can all change the result.

What happens when an AI image is generated?

When someone asks an artificial intelligence system to create an image, the request is processed by powerful computers, usually operating inside a data centre. Those computers perform large numbers of calculations before producing the finished picture.

Using an already trained AI model to create an answer or image is known as inference. Image generation generally requires more computing than a simple text task because the system is constructing detailed visual information rather than producing a sequence of words.

The International Energy Agency, known as the IEA, says different AI tasks can vary enormously in their electricity requirements. Its April 2026 assessment found that more demanding uses can require hundreds or even thousands of times more energy than simple text generation, depending on the model and complexity of the output.

How much electricity does one AI image use?

There is no reliable universal number.

Research published in June 2026 by the United Nations University Institute for Water, Environment and Health, known as UNU INWEH, found that factors including model choice, resolution, output type and other settings can substantially change the environmental footprint of an AI request.

That means a creator producing one finished image is not necessarily using the same amount of electricity as someone repeatedly generating dozens of alternatives. Higher quality or more demanding outputs can also require additional computing.

AI systems are becoming more efficient too. The IEA says improvements in hardware and software have driven down the electricity needed for many individual AI tasks, but total demand can still rise because far more people are using the technology.

Why does water matter?

The computers inside data centres generate heat and need cooling. Some facilities use water as part of that process, while additional water can be connected to electricity generation and the manufacture of the specialist chips used in AI hardware.

Water use varies greatly between facilities. A data centre supplied with lower carbon electricity can still have a significant water footprint because the amount of water required also depends on cooling technology, climate and how the facility itself is designed.

This is why claims that every AI image uses exactly the same amount of water should be treated cautiously. Different models and data centres can produce very different environmental results even when the user appears to be carrying out a similar task.

Is one AI image a major environmental problem?

One image on its own is not the main issue. The more important question is what happens when relatively small individual demands are multiplied across millions of users and businesses.

A production team generating hundreds of concept images for characters, locations or promotional artwork may make considerably more requests than a creator producing only a handful of finished pictures. If that behaviour becomes common across entertainment and other industries, the total amount of computing required can grow quickly.

This is one of the central issues in the wider debate about AI energy consumption. Individual generations can become more efficient while overall electricity demand continues rising because AI is being used more frequently and for increasingly complicated creative work.

What does this mean for film and television?

AI image generation can have useful applications across entertainment. It can help filmmakers explore early ideas, develop visual references, test character concepts or give smaller teams access to creative tools that may previously have been beyond their budgets.

Research published in June 2025 by the British Film Institute, known as the BFI, and CoSTAR found that generative AI was already being explored across the UK screen industries. Their work highlighted creative opportunities while also identifying energy consumption and carbon emissions as issues that need to be considered as adoption increases.

The environmental discussion therefore does not need to become an argument against AI generated art. The more useful question is how these tools are used and whether unnecessary generations, inefficient systems or poorly understood infrastructure increase their footprint.

Could AI image generation become greener?

There are reasons to expect individual generations to become more efficient. Better processors, improved software and more efficient data centres can all reduce the electricity required for a particular task.

Cleaner electricity and better cooling systems can also reduce parts of the environmental footprint. Greater transparency from technology companies would make it easier for creators and businesses to understand the impact of different models rather than relying on broad estimates.

The challenge is scale. If generating images becomes cheaper, faster and easier, people may simply create far more of them, meaning improvements in efficiency do not automatically translate into lower overall environmental demand.

So are AI generated images bad for the environment?

AI generated images have a real environmental footprint, but describing every image as environmentally damaging without context is too simplistic. The impact depends on the model, image settings, data centre, electricity source and how many generations are being produced.

For entertainment, that distinction matters as AI images become part of concept development, digital storytelling and independent production. The technology can create genuine opportunities, but understanding the energy, water and infrastructure behind it gives creators and audiences a clearer picture of what digital creativity actually requires.

For more on how artificial intelligence is changing film, television, music and digital storytelling, explore SoapLandTV’s AI & Entertainment page. entertainment.

Sources: International Energy Agency, April 2026; United Nations University Institute for Water, Environment and Health, June 2026; British Film Institute and CoSTAR, June 2025.