How companies are reducing their carbon footprints

What is a digital carbon footprint?

A digital carbon footprint comprises greenhouse gas emissions (GHGs) produced by information and communications technology, including the following:

• Applications.
• Compute processing.
• Desktop computers.
• Displays and digital signage.
• Laptops.
• Printers.
• Servers.
• Software.
• Smartphones.

The use of IT equipment and software generates emissions through the following methods:

• Their development, manufacturing and usage.
• The energy required to run them.
• Data transfers.
• The energy needed to keep servers and data centers online.

Each component requires energy to be built and then to function. This includes mining, transport and refining of raw materials, transformation and manufacturing, delivery to consumers and end-of-life processes, including recycling and waste incineration, according to The Shift Project, a think tank based in Paris, in its 2025 report, "AI, data, and computing: shaping infrastructures for a decarbonized world."

The Shift Project also noted that the global network infrastructure required for connectivity and data transfer -- including fiber-optic systems, mobile network antennas, and submarine and terrestrial cabling -- contributes to digital carbon footprints, as well.

In the context of ESG, IT leaders struggle to balance their organizations' sustainability goals with maintaining and improving business performance. While some may believe that better performance equates to a larger digital carbon footprint, this isn't always the case, according to Thomas Achhorner, chief digital and information officer at sustainability consultancy ERM, based in London.

For example, when employees have access to more networking performance and capacity, they can be more efficient when working remotely. This, in turn, decreases the need for them to conduct business in person all the time, therefore reducing carbon emissions generated when they commute or travel across regions or countries to meetings.

"It's not only about the footprint created by IT," Achhorner said. "I get a lot more leverage by thinking about how I support the digital footprint in other business areas."

Steps to reduce your digital carbon footprint

IT leaders have various routes they can take to reduce their digital carbon footprint. Some paths are more straightforward than others, but they all lead to the same goal: reducing the business's environmental impact.

1. Decrease IT's Scope 2 impact

The Greenhouse Gas Protocol, a standards body based in Washington, D.C., divides carbon emissions into three main categories:

• Scope 1 comprises all the direct emissions a company generates, such as those produced by company vehicles, manufacturing and processes.
• Scope 2 includes indirect emissions from purchased energy, such as the electricity to power devices, equipment and offices.
• Scope 3 accounts for the indirect emissions produced all the way through an organization's supply chain, including at the supplier and customer levels. This includes raw materials and precious metals used in manufacturing digital devices.

To reduce IT's Scope 2 emissions, organizations can purchase energy-efficient equipment. They could also adopt a circular approach to IT procurement, Achhorner said, which considers the entire lifecycle of devices from manufacturing to their retirement and beyond.

IT leaders can also activate certain settings on devices and networking systems to reduce energy consumption, Achhorner said. This includes programming machines to go into sleep mode after hours.

2. Reduce IT's Scope 3 emissions

Scope 3 is arguably the most difficult category to address, as companies have far less control over how their suppliers and vendors approach ESG.

For example, many cloud computing providers make assertions about their green efforts, but verifying them can be difficult. While IT leaders may initiate conversations with vendors about initiatives to minimize their digital carbon footprint -- such as increasing their use of renewable energy to power data centers -- they sometimes can only rely on claims in the literature, Achhorner said.

That said, organizations can still adopt best practices to reduce Scope 3 emissions in smaller ways. For example, they can encourage employees to send links, rather than attachments, to share files over email.

"If you attach files to emails, that causes a surprising number of emissions," Achhorner said. "If you attach a link to something like SharePoint or a cloud directory, the footprint is significantly less."

3. See where and how AI fits into your strategy

As AI drives energy consumption in data centers, IT leaders must balance using AI for business performance and controlling its negative impact on the environment.

To achieve this, IT leaders can use AI for tasks such as energy modeling in data centers to optimize usage and water consumption, said Abhijit Sunil, a senior analyst at Forrester Research.

Additionally, IT could intentionally pair AI models with specific business functions, Sunil said. For example, a team of editors doesn't have the same needs as a group of software developers. Not everyone requires the latest, greatest and largest version of a tool to be effective in their roles.

It's also helpful to train employees on how to use AI efficiently, which can cut down on energy consumption. An increasing number of companies offer classes so employees can grow more adept at prompting, Sunil said.

"Rather than spend 15 minutes trying to get the right outcome, you optimize your query," Sunil said.

4. Build greener software

Thanks to the adoption of virtualization, containerization and the ability to run more apps on a single server, today's software systems have a smaller carbon footprint than legacy apps, according to Gadhu Sundaram, innovation lead for sustainability technology strategy and advisory at NTT Data and chair of the Green Software Foundation, a non-profit focused on developing standards and tools to build more sustainable systems.

Still, there is room for improvement. The Green Software Foundation prioritizes three areas of focus:

• Using less hardware -- for example, scheduling longer retirement cycles.
• Consuming less energy.
• Being smarter about the energy the organization consumes.

Additionally, IT teams can be more intentional about the apps and features they employ and develop, and evaluate whether they will really bring value to the business.

"The greenest feature is the one you did not develop at all," Sundaram said.

Basic hygiene, too, can go a long way, including the following practices:

• Getting rid of old copies of test data that are no longer relevant.
• Removing apps that no one uses anymore.
• Taking inventory of the company's software and eliminating redundant packages.

Then, the next natural step is to analyze code for optimization.

5. Measure carbon intensity

Organizations can link carbon-intensity measurement to better sustainability outcomes through gamification, Sunil said. This may look like a dashboard that employees can consult to learn how much carbon they emitted in a given period, and what that means for the planet.

"For many employees, the carbon they emitted from an AI query may not make any difference [to them], but perhaps a metric like, 'You saved X number of trees this month' [would]," Sunil said. "It may be a gamified way in which behavioral changes can be brought about."

6. Embed sustainability into IT governance

To ensure sustainability becomes part of IT's culture, leaders must establish metrics that people can use to measure and monitor progress.

Some tasks and tools to measure include the following:

• AI inferencing.
• AI modeling.
• AI workloads.
• Carbon intensity.
• Cloud efficiency.
• Power usage effectiveness.

"Then we could make broader business connections, like the cost savings gained from our energy efficiency initiatives overall," Sunil said.

Presumably, companies will use sustainability management software to gather this data. However, IT leaders should push vendors to be transparent about the digital carbon footprint of their own services -- especially if they are powered by AI, Sunil said.

"Right now, we are not really emphasizing the amount of carbon footprint that the AI solution will, in itself, produce," Sunil said. "On the other hand, some of these AI [features] may not even be needed for certain capabilities, but [vendors] have to talk about it because AI has become a differentiator for many of these companies."

Carolyn Heinze is a Paris-based freelance writer. She covers several technology and business areas, including HR software and sustainability.