Can big tech code its way out of carbon?

Artificial intelligence is rewriting the rules of productivity, business, and global infrastructure. But while AI unlocks new frontiers in efficiency and creativity, it’s also triggering a wave of energy demand that few are prepared for and one that could redefine who the world’s biggest polluters are.

For tech companies, which have long pledged to rapidly draw down their planet-warming emissions, the AI revolution has disrupted their climate ambitions. Just a few years ago, many saw the climate pledges from tech companies as entirely achievable. Today, those same companies are entering a phase of deep uncertainty one in which they no longer have a clear picture of where their emissions will land, especially with the rise of AI and the exploding energy demand from data centers.

In this piece, I explore what’s driving Big Tech’s growing interest in carbon removals, how it connects to the future of AI, and what it tells us about the future of the carbon market.

Ex-Google CEO Eric Schmidt says AI could one day use 99% of the world's electricity right now it’s just 3%—but with bigger, faster models, it could skyrocket.

What many don’t realize is just how energy-hungry AI already is. A single ChatGPT query can use up to ten times more electricity than a standard Google search. That difference may seem small in isolation, but when you multiply it by billions of interactions a day, the emissions start to add up fast.

The International Energy Agency estimates that data center electricity consumption could more than double by 2030, reaching nearly 1,000 terawatt-hours—more than the entire national grid of Japan. In the U.S., data centers could account for 12% of total electricity demand by the end of this decade, up from just 3% in 2022. And much of that growth will be driven by AI.

Microsoft’s greenhouse gas emissions were 29.1% higher than 2020 levels last year, amid growing demand for energy to support artificial intelligence. It last reported annual GHG emissions of 17.2 million tons at the end of 2023.

Even as tech companies accelerate investments in renewable energy, residual emissions remain—especially in Scope 3: the outsourced compute, cloud platforms, hardware supply chains, and energy consumed by users running AI tools globally. This is where things get complicated, and where Big Tech is beginning to feel real pressure.

One of the clearest signals came recently from Sam Altman, who told the U.S. Senate:
“Eventually, the cost of intelligence, the cost of AI, will converge to the cost of energy. In terms of long-term strategic investment for the US to make, I can't think of anything more important than energy.”

It’s a concise but powerful formula:
AI = energy.
And energy, unless fully decarbonized, = emissions.

The key metric to understand here is called carbon intensity, which is basically a measure of how many grams of carbon dioxide pollution are released for every kilowatt-hour of electricity that’s produced.

Carbon Intensity and Electricity Mix in Big Tech Data Center Countries

Country Carbon Intensity (gCO₂e/kWh)
United States 400–500
Ireland 300–400
Sweden 50–100
France 50–80
Germany 300–400
Denmark 100–150
China 600–700
Netherlands 300–400

SOURCE: Yearly Electricity Data (2025); Energy Institute – Statistical Review of World Energy (2024) Our world in Data

One of the biggest blind spots in today’s AI-driven emissions conversation is the lack of public data. As MIT Technology Review reports, the vast majority of tech giants still don’t disclose how much electricity their AI workloads actually consume. While Microsoft, Google, and other major players release broad sustainability reports, they rarely provide detailed data on the emissions or energy use specifically tied to AI. This lack of granularity makes it extremely difficult for regulators, investors, and researchers to understand the true climate impact of AI at scale. But this isn’t just a gap in reporting—it’s a strategic blind spot. As AI becomes core to their business models, the absence of transparency limits accountability and hinders efforts to develop effective standards for managing its environmental footprint.

What’s happening now is that the tech companies leading the AI boom—Microsoft, Google, Amazon, Meta—are becoming increasingly aware that, if unchecked, their emissions footprints could rival or even surpass those of fossil fuel majors. This isn’t about a PR risk anymore. It’s about existential exposure to regulation, investor pressure, and public backlash. They are deeply concerned about what this means for their growth, reputation, and license to operate.

That’s why these companies are moving fast into one of the most experimental and underdeveloped parts of the climate solution set: carbon removals. Carbon removals are gaining traction rapidly, thanks to the boom in net zero targets—both at national and corporate levels. To truly “net” emissions, companies must remove their residual emissions using either natural removals or technological solutions.

Today, Microsoft and a small group of peers represent more than 90% of global demand for high-durability carbon removals. Initially, many in the carbon market assumed this was performative—a way for wealthy firms to position themselves as pioneers while buying expensive credits for branding and communication purposes. But that reading misses the strategic logic. Big Tech is not just buying tons. It’s buying time, optionality, innovation, and influence over a market that could become essential to its future.

The top ten CDR buyers:
Microsoft: 20,377,120 tons
Frontier Buyers: 1,249,429 tons
Google: 563,417 tons
Airbus: 400,000 tons
Equinor: 330,000 tons
Amazon: 250,000 tons
NextGen CDR: 212,000 tons
SkiesFifty: 200,000 tons
BCG: 195,804 tons
Stripe: 176,136 tons

Source: cdr.fyi

Carbon removals—especially technologies like direct air capture, biochar, enhanced weathering, and nature-based removals—are increasingly seen as critical infrastructure. They give companies a way to compensate for emissions that can’t be abated in the short term. They also offer regulatory flexibility and a hedge against future constraints. If carbon disclosure frameworks like the SBTi, the SEC’s climate rules, or Europe’s CSRD become stricter, having access to a portfolio of high-integrity removals is a serious advantage.

They’re not just buyers—they’re builders. Microsoft has signed multi-million, long-term offtake deals with companies like Exomad Green and 1PointFive. This year, the company signed several major CDR deals, including three with BECCS developers CO280, AtmosClear, and Stockholm Exergi—expected to capture and store over 10 million tonnes of CO₂ over the next 15 years. Meta and Stripe are behind Frontier, a billion-dollar advance market commitment to scale up early-stage carbon removal. Google is blending removals into its broader energy and sustainability strategy. Amazon is backing carbon-focused startups through the Climate Pledge Fund.

But this isn’t just about offsetting emissions. These companies are using the voluntary carbon market (VCM) for what it was originally meant to do: drive innovation and unlock finance for entrepreneurs building the next wave of climate tech. By signing advance contracts and paying high prices today, they’re de-risking R&D, creating demand signals, and making it possible for entirely new carbon removal technologies—and business models—to emerge.

And they’re not only focused on tech-based removals. Last year, four of the biggest technology companies—Google, Meta, Microsoft, and Salesforce—announced the formation of the Symbiosis Coalition, a new joint effort to purchase 20 million nature-based removal credits by 2030. This approach aims to support both climate impact and ecosystem integrity while scaling CDR supply.

For Big Tech, this is not philanthropy. It’s market-making. It’s about shaping the infrastructure they know they’ll rely on for decades to come.

There’s a deeper reason for all of this. Big Tech understands that as AI reshapes everything, it must also decouple progress from pollution. Without credible ways to address its footprint, the sector could face growing scrutiny and political blowback. Carbon removals are not a license to pollute, they’re a climate insurance policy while the clean energy transition struggles to keep pace with the speed of digital growth.

This isn’t a theoretical risk. It’s a real, near-term challenge. And the removals market still immature, expensive, and supply-constrained may be the only lever that can scale fast enough to match the trajectory of AI’s emissions curve.

So while governments debate climate policy and legacy sectors drag their feet, the tech companies driving the next industrial era are quietly rewriting the rules. They’re not just betting on intelligence. They’re betting on carbon. And they know the two are more connected than most people realize.

This article is also published on Substack. illuminem Voices is a democratic space presenting the thoughts and opinions of leading Sustainability & Energy writers, their opinions do not necessarily represent those of illuminem.