There’s a particular kind of irony in 2026: we’re building machines that can write poems, diagnose bugs, and summarize our lives in seconds, while the electricity needed to power all that “magic” keeps dragging the climate conversation back to Earth. So when I saw that Anthropic has joined Frontier—the carbon removal buyers’ collective—and helped push a new $915 million funding package, I read it less as a routine corporate pledge and more like a signal flare from the AI world.
Anthropic’s entry matters because it’s not just another tech company showing up with a check. It’s described as the first AI startup to join Frontier. In a moment when AI firms are openly scrambling for power—sometimes clean, sometimes not—this looks like an attempt to take a more deliberate position in the climate ledger.
A bigger war chest, and a clearer message
The new commitments nearly double Frontier’s total, taking pledged funding to $1.8 billion. That’s not theoretical money either. Frontier has already signed almost $700 million in contracts across more than 50 projects, backing the removal of about 1.8 million tons of carbon.
If you’re wondering what these companies actually “get” out of it, the answer is straightforward: buyers typically use Frontier-backed carbon removal credits to reduce their publicly reported carbon footprints. Think of it like financial accounting, but for emissions—credits can be applied to help offset pollution that companies still can’t eliminate in day-to-day operations.
And let’s be honest: that’s exactly where the moral tension lives. Carbon removal credits can represent real climate value, but they can also become a permission slip to keep emitting. The difference comes down to quality, verification, and whether the buyer is reducing emissions at the same time.
Why Anthropic’s move is louder than it looks
Frontier has had major names in the mix since the beginning—Google is a founding member, alongside Stripe and Shopify. But Anthropic arriving as the first “pure” AI startup in the group lands differently because AI’s energy appetite is no longer a footnote. It’s front-page math.
This is also Anthropic’s first climate-related deal of this kind. The company hasn’t released a sustainability report yet, and it has previously described its energy posture as “all of the above”—a phrase that, in practice, often reads as: we’ll buy whatever power we can secure, including electricity that’s far from clean.
So is this a genuine shift, or a strategic hedge? I can’t see inside the boardroom, but the timing is hard to ignore. Joining a collective that is explicitly about scaling carbon removal looks like a step toward being taken seriously on climate—especially when scrutiny around AI infrastructure is only getting harsher.
Frontier’s role: a shared gatekeeper for carbon removal
Frontier was built by tech companies to help meet climate promises in a world where “just stop emitting” isn’t immediately possible. Even the most ambitious corporate climate plans run into stubborn categories of pollution—aviation is the classic example—where solutions are expensive, immature, or simply not available at scale yet.
At the same time, carbon removal is still a young industry. There aren’t many large, mature suppliers that can reliably remove enormous volumes of CO2 today, let alone at prices companies want to pay. Frontier’s model is to evaluate carbon removal providers and sign contracts with the ones it believes can actually deliver.
In other words, Frontier functions as a kind of collective due diligence engine—a shared resource for companies that want to buy carbon removal without individually reinventing the verification process each time.
Less scattershot, more “can this reach a gigaton?”
With this latest funding announcement, Frontier is also changing its posture. It says future funding will come with higher scrutiny. Practically, that means fewer deals, tighter standards, and more focus on approaches that could plausibly scale to removing one gigaton (that’s 1 billion metric tons) of CO2 per year—or more.
New contracts are expected to run about eight to 10 years, which is a tell in itself. Carbon removal isn’t something you “pilot” in a quarter and declare victory. These are long arcs—engineering, permitting, monitoring, financing—so longer contract durations can be the difference between a clever prototype and a bankable project.
This shift—from many smaller bets to fewer, bigger ones—echoes what we’ve seen elsewhere in the market, including the pattern associated with Microsoft, which has been the largest buyer of carbon removal credits. The industry seems to be moving from curiosity to consolidation: pick the potential winners and help them scale fast.
What Frontier has funded so far
Since launching in 2022, Frontier has backed a mix of carbon removal pathways, including:
- Direct air capture (pulling CO2 straight from ambient air)
- Enhanced rock weathering (accelerating natural mineral processes that bind CO2)
- Bio-oil approaches (turning biomass into stable carbon-rich liquids for storage)
- Ocean “anti-acidification” concepts (methods aimed at shifting ocean chemistry to store more carbon)
- Bioenergy with carbon removal and storage (energy generation paired with capturing and storing CO2)
If that list sounds like a mix of science, industry, and a little bit of sci-fi, that’s because carbon removal is still in the “many tools, uncertain winners” phase. Frontier’s new emphasis suggests it wants to narrow the field toward methods that don’t just work—but can scale massively.
The uncomfortable question: who pays, and for how long?
Here’s the part that always gets glossed over when climate pledges make headlines: companies may want this market to mature, but they’re also signaling they don’t intend to subsidize it forever.
Frontier has said that for every new contract, the carbon removal company needs to show a credible pathway to government subsidy or support. That’s not a small statement—it’s an admission that voluntary corporate buying won’t be enough to build gigaton-scale carbon removal on its own.
The UN’s IPCC has repeatedly said carbon dioxide removal will likely be necessary if the world is going to reach net-zero. Yet demand is thin relative to the scale required, partly because the benefits are shared globally while the costs are paid locally—an old public-goods problem. Like clean water infrastructure, the bill tends to end up with governments.
Frontier has indicated it plans to keep signing contracts through 2040. If you’re reading that as both ambition and a countdown clock, you’re not wrong. The subtext is clear: carbon removal is becoming a serious industrial category—but the handoff from corporate early adopters to public funding may be the only way it ever reaches the scale the climate math demands.
Comments