
Anthropic, the artificial intelligence startup behind the Claude family of models, has quietly taken a major step into climate action by joining Frontier, the coalition dedicated to advancing carbon removal technologies. This makes Anthropic the first pure AI startup to become a member of the group, signaling a potential shift in how the rapidly growing AI industry approaches its environmental footprint.
The company is contributing to a new $915 million tranche of funding for Frontier, which nearly doubles total pledged funds to $1.8 billion since the coalition's founding in 2022. So far, Frontier has contracted approximately $700 million across more than 50 projects, achieving the removal of 1.8 million metric tons of carbon dioxide. Companies that pledge money to Frontier typically use the resulting carbon removal credits to offset their publicly reported emissions, allowing them to continue emitting while funding technologies that pull CO2 from the atmosphere.
Anthropic's decision comes at a time when the AI sector is under increasing scrutiny for its energy consumption. Training large language models and powering inference servers requires massive amounts of electricity, and many data centers still rely on fossil fuels. While some tech giants like Google and Microsoft have made aggressive climate commitments, the industry as a whole has been criticized for opaque reporting and a tendency to favor expedient energy sources over truly clean ones. Anthropic itself has not yet published a sustainability report and has previously stated it supports an "all of the above" approach to energy, a phrase that often translates into purchasing power from a mix of sources, including natural gas and coal.
The Carbon Removal Landscape
Frontier was founded by Stripe, Google, Shopify, and other technology companies to address a fundamental problem: many companies want to reach net-zero emissions within the next one to two decades, but certain emissions — such as those from air travel, heavy industry, and agriculture — are currently impossible to eliminate entirely. Carbon removal offers a way to compensate for these residual emissions by extracting CO2 that has already been released into the atmosphere and storing it permanently underground or in durable materials. However, when Frontier launched, the carbon removal industry was still in its infancy. Few companies could demonstrate commercial-scale operations, and the cost per ton of CO2 removed was prohibitively high.
Frontier acts as a kind of venture capital fund for carbon removal, vetting startups and signing offtake agreements — contracts to purchase future credits — for those it believes have the best chance of scaling. By pooling demand from multiple buyers, Frontier reduces transaction costs and provides developers with the financial certainty they need to build pilot plants and eventually full-scale facilities. Over the past two years, Frontier has backed a diverse portfolio of technologies, including direct air capture (DAC), enhanced rock weathering, bio-oil injection, ocean alkalinity enhancement, and bioenergy with carbon capture and storage (BECCS). Each approach has its own trade-offs in terms of cost, permanence, scalability, and environmental impact.
With the new $915 million tranche, Frontier is also refining its strategy. The coalition announced that future funding will come with a higher level of scrutiny. Instead of spreading smaller investments across many early-stage projects, Frontier plans to concentrate on fewer, larger offtake agreements with companies that demonstrate a clear path to removing at least one gigaton — one billion metric tons — of CO2 per year by mid-century. New contracts will typically span eight to ten years, with some extending as far as 2040. This shift mirrors a similar trend at Microsoft, which has been the largest single buyer of carbon removal credits and recently began consolidating its purchases around a smaller number of high-volume suppliers.
Frontier is also making it clear that it does not intend to underwrite the carbon removal market indefinitely. For any new contract, the carbon removal company must "show a path to government subsidy/support," a Frontier spokesperson told TechCrunch. This reflects a growing awareness that, like clean water or public health, climate stabilization will ultimately require sustained government investment. The UN Intergovernmental Panel on Climate Change (IPCC) has repeatedly stated that carbon dioxide removal is essential for achieving net-zero emissions, yet neither consumers nor companies have shown much appetite for bearing the full cost. Carbon removal today still costs between $100 and $600 per ton, depending on the technology, and demand is largely driven by voluntary corporate pledges rather than regulatory mandates.
AI's Energy Paradox
Anthropic's entry into Frontier is notable because it highlights the tension at the heart of the AI industry. AI companies are on an energy buying spree, signing power purchase agreements with solar and wind farms, but also turning to natural gas and even exploring small modular nuclear reactors. The sector's voracious appetite for compute has led to a resurgence in fossil fuel infrastructure in some regions, as grid operators scramble to support new data centers. Critics argue that without strong climate guardrails, AI could become a net negative for the environment, locking in decades of emissions even as it drives efficiency gains elsewhere.
By joining Frontier, Anthropic is signaling that its leadership recognizes the need to address residual emissions — a step that could pressure other AI startups to follow suit. Until now, AI companies have largely focused on operational efficiency, such as using more efficient chips or locating data centers near renewable sources. But direct carbon removal offers a way to tackle the hardest-to-abate portion of their footprint. For Anthropic, which has yet to publish its own carbon footprint or sustainability roadmap, the Frontier membership could be the beginning of a more comprehensive climate strategy.
The move also positions Anthropic among a growing number of tech firms that view carbon removal as a necessary complement to renewable energy and energy efficiency. Google, a founding member of Frontier, has already committed to purchasing carbon removal credits to offset its entire historical emissions. Stripe has integrated carbon removal into its business model, using a portion of transaction revenue to fund removal projects. Shopify has been an early backer of several carbon removal startups. Anthropic, by entering the coalition, can now tap into Frontier's expertise and the collective purchasing power of its members, potentially accelerating its own climate efforts.
Key Facts
- Anthropic is the first pure AI startup to join Frontier, contributing to a new $915 million funding tranche.
- Total pledges to Frontier now stand at $1.8 billion, up from approximately $885 million previously.
- Frontier has contracted nearly $700 million across more than 50 projects, removing 1.8 million metric tons of carbon.
- Anthropic has not yet published a sustainability report; this is its first climate-related deal.
- Frontier will now focus on fewer, larger offtake agreements, with contracts lasting 8–10 years and extending to 2040.
- Future contracts require carbon removal companies to demonstrate a path to government subsidy or support.
- Technologies funded include direct air capture, enhanced rock weathering, bio-oil, ocean alkalinity enhancement, and BECCS.
- Frontier was founded by Stripe, Google, and Shopify to pool demand for carbon removal.
Since its launch, Frontier's approach has evolved. Early investments were scattered across many novel technologies, many of which failed to scale. The new discipline around larger, longer-term contracts aims to accelerate the few most promising approaches toward the gigaton scale. Frontier is essentially acting as a market maker, providing the demand signal that the carbon removal industry needs to attract further private and public investment. If successful, the coalition could help bring down costs to the point where carbon removal becomes a viable tool for reaching global climate targets.
However, the road ahead is uncertain. Carbon removal remains a controversial topic within environmental circles. Some activists argue that it risks becoming a "moral hazard," allowing polluters to continue emitting under the guise of future removal. Others worry that focusing on removal diverts attention and resources from emission reductions, which are more immediate and often cheaper. Frontier and its members counter that reductions alone will not be enough; the IPCC and the International Energy Agency have both concluded that the world will need to remove billions of tons of CO2 annually by mid-century. Waiting to develop the technology until after emissions have peaked would be too late.
As for what happens after 2040, Frontier has not specified. The coalition seems to hope that by then, governments will have begun to take over, either through carbon pricing, direct procurement, or public utilities. If governments do not step in, the burden will fall on voluntary corporate buyers, but that model may be insufficient to reach the scales required. The climate is warming at an accelerating rate, and the window to deploy carbon removal at scale is narrowing. Anthropic's decision to join Frontier may be small in the broader scheme of global emissions, but it represents a noteworthy shift in the AI industry's coming of age when it comes to climate responsibility.
Source:TechCrunch News
