How Mastercard is trying to tame its digital carbon footprint

Companies in every sector are investing in artificial intelligence and digital services to create new business value, spurring hundreds of billions of dollars of spending on data center expansion projects by the biggest cloud-services players and co-location providers. 

Those investments will derail corporate emissions goals if they’re not managed properly. Getting ahead of that outcome will take closer collaboration between chief technology or information officers and sustainability leaders.  

“I know it’s not easy to carve out bandwidth to focus on this problem, but I would encourage all of my peers to do this,” said George Maddaloni, chief technology officer, operations at Mastercard.

Mastercard stepped up efforts to more closely manage its digital carbon footprint three years ago, before AI strategy was top of mind for every business executive. In April, the company formally made environmental sustainability one of the key performance indicators reviewed monthly by a new steering committee composed of senior executives, including CSO Ellen Jackowski.

“We started with the data about what we run from a technology perspective, and then looked at how to allocate and assign a footprint to our different products and services based on that,” Maddaloni said.

Growing concern

Information technology accounts for an estimated 2-4 percent of annual global emissions, according to the International Energy Agency — a figure that’s growing rapidly as companies increasingly rely on digital services — with hardware manufacturing, life-cycle management and electricity accounting for the biggest chunks of the total. 

Large tech companies with aggressive emissions reduction goals, including Amazon, Google and Microsoft, are struggling to achieve those targets in large part because of the more than $364 billion they spent this year on data center buildouts.

The potential ripple effect affects sustainability professionals beyond Big Tech. More than 60 percent of the leaders surveyed in August by The Conference Board indicated that data center energy demand was their greatest concern related to their company’s AI investments, followed by the emissions related to the services themselves.

The level of emissions generated by IT infrastructure, including data centers, varies widely from industry to industry. Enterprise technology contributes an estimated 60-65 percent of emissions related to electricity (Scope 2) and upstream and downstream business activities (Scope 3) at banks and financial services firms. For healthcare providers, the average is closer to 10-15 percent of Scope 2 and 3.

Mastercard’s data center footprint represents 60 percent of emissions from its direct operations (Scope 1) and purchased electricity. Those two categories account for 10 percent of its total emissions, which means Mastercard’s data centers contribute 6 percent of the company’s entire carbon footprint.

“This is a material topic for a handful of industries, including financial services,” said Bjoern Stengel, global sustainability practice lead at tech research firm IDC. “With the rise of AI, it became a mainstream topic overnight.”

Best practices for taming digital footprints

Over the past three years, Mastercard has managed to decouple its growth in its payment services from its emissions. In 2024, for example, the company’s revenue grew 12 percent, but Mastercard’s overall emissions decreased 7 percent.

Key to that achievement was the creation of a patent-pending management dashboard that includes real-time electricity consumption of Mastercard’s services (including the percentage that comes from renewables), information about server and hardware use and carbon-intensity metrics at the product, program and asset level.

The information is used to generate scores that the committee and division heads can use to compare and evaluate the impact of various decarbonization efforts. Among the metrics considered is the carbon intensity of the electric grid where a data center is located. 

Here are some specific tactics that have helped with Mastercard’s IT transformation:

• Color-coding for dashboard scores: Teams can quickly see how their product or service is performing; red means that an initiative falls in the bottom third for energy intensity, renewable use and hardware efficiency.
• Carbon profiles for each product or service: This includes customer-facing and internal assets. Product leads are responsible for understanding and managing energy consumption.
• Proactive decommissioning: Mastercard removed more than 1,200 computer servers in 2024, consolidating the jobs they handled, and retired technology that wasn’t being used.
• Closer scrutiny of cloud services and co-location partners: Mastercard has used that information, in some cases, to switch where specific services are running based on the carbon intensity of certain regions. It uses actual data from these suppliers, rather than spend-based estimates.
• Emissions-sensitive software code: Mastercard is being judicious about the data chosen to train AI models, which keeps them smaller and saves energy.

Offense and defense

Sustainability leaders are helping the most mature organizations, such as Mastercard, both to manage the environmental impact of AI and to brainstorm ways it can be used to advance business value.

“Instead of looking at this as a siloed topic, look at this as part of the bigger AI ROI equation,” said IDC’s Stengel. “There are financial and nonfinancial sides to this discussion.”

AI enables companies to use information that’s been collected for ESG reports and greenhouse gas inventories for much more than compliance, said Sammy Lakshmann, a U.S. PwC partner focused on digital and AI-enabled sustainability strategy.

For example, data that the sustainability teams collect about extended producer responsibility laws offer important signals for product managers and finance teams about potential future fees, he said.

Likewise, real-time climate data could be used by retailers to adjust merchandising strategies or product inventories proactively. 

“The companies that are going to win are the ones that combine AI, sustainability and business value,” Lakshmann said.