Are AI Videos Bad for the Environment? Here Is the Actual Data

Large film productions can generate as much as 2,840 tonnes of CO2. Transport alone accounts for 51% of emissions, followed by electricity and gas at 34% and diesel generators at 15%.

AI video production can cut down on many of these emissions, especially for projects that would otherwise require large crews, international travel, and extensive equipment. But AI comes with its own environmental costs.

Every AI-generated video uses computing power and electricity. And with millions of people creating AI videos, whether for major productions or just for fun, the combined environmental impact is worth paying attention to.

So, the question isn't simply whether AI is better for the environment, but how we use it and what we create.

We'll look at AI's environmental impact, its water and electricity consumption, and how its carbon footprint compares to traditional film production.

AI videos have an environmental impact, but we also need to consider what they're replacing. For large-scale productions, AI can help reduce several sources of carbon emissions.

Here's what AI video production can help avoid:

• International and local travel and transportation, which contribute significantly to carbon emissions.
• On-site infrastructure, such as generators, and construction materials used for physical sets.
• Crew and cast accommodation, which requires energy for lighting and laundry.
• Food waste and prop waste generated during production.
• Thousands of plastic water bottles and paper plates used on busy filming days.

However, generating AI videos also consumes resources. Producing large volumes of content, even for casual entertainment, adds to that environmental impact. That's why how we use AI matters.

AI does affect the environment. AI computing can use 10 times more energy than conventional computing tasks, putting greater demand on electricity, water, and natural resources.

But traditional filmmaking comes with its own environmental costs. Major US film productions reportedly emit 70% more greenhouse gases than the average US business. These gases, including CO₂, methane, and nitrous oxide, contribute to global warming, rising sea levels, and damage to ecosystems.

Now consider a major commercial that would normally require hundreds of crew members, international flights, generators, and elaborate sets. AI can replace some of that physical production, cutting travel, fuel use, and waste.

That's where AI has the potential to make a difference. But generating millions of short videos just because it's easy also consumes resources. AI isn't automatically better for the environment. Its impact depends on what it replaces and the resources it consumes.

Behind the AI-generated videos are data centers equipped with advanced hardware, such as GPUs, to process datasets quickly and efficiently.

In the process, they generate a lot of heat, requiring constant cooling. Additionally, data centers require robust power infrastructure for continuous, reliable operations.

Therefore, if the right environmental safeguards are not put in place, AI could affect the environment in several ways:

• Water Consumption: AI data centers need gallons of water to cool their systems. However, to avoid water wastage, most data centers are built in cold-climate areas. AI companies have also embraced liquid and closed-loop cooling systems, significantly reducing water use.
• Energy Use: Training large language models can be energy-consuming. For this reason, tech companies with massive data centers like Meta, Amazon, and Google are transitioning to renewable energy, making them the leading corporate buyers of clean energy. Therefore, adopting AI for video production reduces the carbon footprint in the long run.
• E-Waste: The chips and the hardware in data centers eventually become electronic waste. However, the chips are very valuable and are not easily discarded. In fact, most outdated hardware in data centers is refurbished and recycled, reducing global demand for mining raw materials and preventing soil erosion.

The Real Data on AI Water Usage and Energy Consumption

Let us look at some real statistics about AI water usage and energy consumption:

• Microsoft reduced its water use effectiveness from 2.3 L/kWh in the early 2000s to 0.27 L/kWh in 2025.
• 70% of the leading tech companies that use AI rely on renewable energy and are targeting 100% reliance on it.
• Amazon Web Services expanded its use of recycled water in data centers in 2025 from 24 to over 120 locations in the US.
• Switch Datacenters relies on 100% green energy in all its locations.

Now that we've looked at AI's environmental impact, let's compare it to traditional film production, especially for large-scale projects with hundreds of crew members, extensive travel, and heavy equipment.

Here's what a hypothetical 400-person crew filming a production like the UFC Battle in the Ocean could involve:

Transport & Travel

Travel and transportation of equipment, cast, and crew to different shooting locations account for 40–60% of a production's carbon footprint. A 400-person crew could require hundreds of flight tickets, road transport to and from the airport, and the movement of production equipment.

Transporting heavy gear, like cranes needed for a physical production of a video like UFC Battle in the Ocean, by air consumes a massive amount of fuel.

On the other hand, using AI to create a cinematic video like UFC Battle in the Ocean means smaller crews, less travel, and little need to transport heavy production equipment. This can eliminate much of the aviation and freight emissions associated with a large physical shoot, although generating the video digitally still requires energy.

Onsite Generators & Hotel Energy

A lot of large crew productions take place in remote locations without sufficient power access. If UFC Battle in the Ocean were produced in person, generators might be needed to power lighting, cameras, cranes, and sound equipment.

Diesel fuel produces about 2.68 kg of CO2 per liter. With a generator operating around the clock during physical film production, the amount of CO2 produced in a day can rise considerably.

In addition, accommodating such a large team increases energy consumption through lighting, HVAC systems, and laundry, and could account for at least 6% of the production's carbon footprint.

These physical production requirements weren't needed on the same scale while creating UFC Battle in the Ocean. By using AI, we avoided much of the infrastructure and energy consumption that a comparable physical shoot could have required. The digital production still consumed electricity, but it removed several major sources of emissions associated with large-scale filming.

Catering Logistics

Catering for a 400-person crew three times a day means preparing around 1,200 meals. Any food that goes to waste can release methane as it decomposes, adding to the production's greenhouse gas emissions.

Then there are the single-use plates, cups, and water bottles, some of which can take up to 500 years to decompose. Transporting food and supplies to the filming location also adds to fuel consumption.

With AI, there's no need to organize catering on this scale or ship food and supplies to a remote set. AI still uses energy, but for a production involving hundreds of people, avoiding these logistics can reduce its environmental footprint.

This is especially relevant to the work we do at THE BLUE HORIZON. We produce cinematic commercials, films, and high-impact visual content for major brands and entertainment projects. These productions can involve elaborate environments, complex scenes, and significant logistical requirements when filmed physically.

We use AI alongside professional filmmaking and creative direction to produce large-scale commercials and cinematic films. For projects that would otherwise require extensive travel, large crews, and elaborate sets, we can cut down on the physical resources involved and the emissions that come with them.

If you have some ideas you’d like to make real, book a creative strategy call with our team.

With all the negative environmental impact we have seen from traditional film production, one would expect it to be easy to hold them accountable. However, they often get a free pass on their carbon emissions while generative AI carries the most blame.

But why is it so easy for traditional film production’s carbon footprint to go unnoticed?

Here are some reasons:

• Behind-the-Scenes Impacts: Traditional filmmaking's environmental impacts happen behind the scenes and aren't immediately visible. Consider the pre-production logistics, including travel, equipment freight, and prop production. These logistics emit carbon even before the production process begins. However, if nobody is tracking the emissions, it’s easily ignored.
• Divided Responsibility: Traditional film production involves multiple entities, each of which affects the environment in some way. For instance, transport companies emit carbon while shipping equipment, while catering companies generate food-related waste. Decentralized accountability makes it hard to track and hold the involved parties accountable.
• Mobile Nature of Traditional Film Production: Film productions often move from one location to another, with crews and equipment traveling between cities or even countries. Keeping track of fuel consumption, electricity use, and waste across different locations can be difficult.
• Lack of Environmental Transparency: Some production companies track and report their emissions, but many don't publish this information. Without clear reporting, it's difficult to know how much carbon a production generates or where those emissions come from.

As the global carbon footprint continues to increase, brands and production companies have more reason to consider the emissions behind the content they create.

For large-scale commercials and films, AI can reduce travel, equipment transport, and other resources needed for physical shoots. But the way a studio uses AI matters too.

When hiring an AI film studio, consider these factors:

• Carbon Accountability: Ask the studio how it measures emissions across its projects and what it's doing to keep them down. It should be able to explain where those emissions come from and how it accounts for them.
• Hosting Provider: Find out which cloud providers the studio uses. Companies such as Microsoft, AWS, and Google are among the leading buyers of renewable energy, although their data centers still consume electricity and other resources.
• Water Usage Effectiveness (WUE): Cooling systems can use significant amounts of water. Some cloud providers have reduced their water consumption through more efficient cooling technologies. For example, AWS has reported a WUE of around 0.15 L/kWh. Ask the studio what information it can provide about the water usage of its infrastructure.
• Hardware Waste Management: What happens to the hardware in the data centers once it's outdated? Some providers refurbish, reuse, or recycle equipment to reduce e-waste and demand for new materials.

Let us answer some commonly asked questions about AI and the environment:

Is Generative AI Bad for the Environment or Just Poorly Understood?

Generative AI consumes electricity, water, and computing resources, so concerns about its environmental impact are justified.

However, when AI replaces a large film production involving extensive travel, heavy equipment, physical sets, and hundreds of crew members, it can help reduce emissions. The problem comes when we treat AI generation as unlimited and create endless videos without considering the resources they consume.

What Happens to the Hardware AI Studios Use After a Project?

AI studios typically use hardware provided by cloud platforms, which run multiple projects on the same infrastructure.

Once the equipment becomes outdated, providers may refurbish, reuse, or recycle its components. However, manufacturing and replacing this hardware still consumes resources and creates electronic waste.

Do Governments Regulate the Carbon Footprint of AI Models?

Yes, but no single standard exists for measuring AI emissions. Regulations vary by country. For example:

• EU: The AI Act requires general-purpose AI model providers to document their energy consumption.
• US: Data centers may need environmental permits for activities affecting air and water, depending on where they operate.

Can You Scale AI Video Production Without Increasing Your Carbon Footprint?

Scaling AI video production generally increases energy consumption, even with efficient workflows.

However, for major commercials and films, AI can replace resource-heavy physical shoots. The key is to use it with purpose, not to generate endless content simply because it's possible.

The answer to the question, ‘Is AI bad for the environment?’ depends partly on the available alternatives.

A major film or commercial shoot can involve hundreds of crew members, international travel, heavy equipment, and elaborate sets. AI can reduce the need for some of these, although it still consumes energy and water.

Using AI for a large production is also very different from generating millions of videos just because we can. Both have an environmental impact, but the scale and purpose of the work make a difference.

THE BLUE HORIZON supports brands, artists, and the entertainment industry in bringing their ideas to life through cinematic videos that blend creative direction with cutting-edge AI technology.

The results? Hollywood-level videos that go viral, get people talking, and drive measurable business results.

Book a call to discuss your project with our team.