INVESTIGATION: Investors poured millions into Earthbanc. Now it's carbon project is being shutdown amidst legal battle

In December 2023, amidst the COP28 climate conference in Dubai, Swedish-British pharmaceutical giant AstraZeneca unveiled its plans to invest a whopping $400mn in tree-planting projects around the world.

Pascal Soriot, CEO of AstraZeneca, told the Guardian at the time that the company aimed to plant 200 million trees by 2030, an effort it estimated would remove around 30 million tonnes of carbon dioxide from the atmosphere – making the pharmaceutical giant “carbon negative”.

As part of that push, AstraZeneca committed to backing the first phase of Regeneration Kenya – an initiative launched by Stockholm-based startup Earthbanc and local agroforestry NGO Green Planet Initiative 2050 Foundation (GPI) in 2022.

According to the project's design documents, the programme aimed to establish up to 35 million trees across 26,000 hectares of farmland in western Kenya, enlisting around 5,000 local farmers who were promised future income through the sale of carbon credits to companies seeking to offset their emissions. AstraZeneca's contribution would fund the planting of six million trees.

Armed with fresh capital, Earthbanc, GPI, and hundreds of local farmers planted approximately half a million trees in 2024 in and around Uasin Gishu county, situated along the spine of Africa's Rift Valley.

But late last year the scheme was abruptly closed down, according to multiple sources familiar with the matter and confirmed by an internal letter addressed from AstraZeneca to the Kenyan Forest Service (KFS).

The view from Mau Summit, Kenya, close to Uasin Gishu county. Credit: Jacek_Sopotnicki/Getty Images/Canva

“Unfortunately, despite the commitment of all, a decision was taken to wind down the project,” the letter reads.

The closure comes amid an ongoing lawsuit brought by GPI, which alleges Earthbanc secretly set up a subsidiary to cut the organisation out of the project's profits and sideline the local community, allegations the startup denies.

However, documents and recordings obtained by Impact Loop along with interviews with several former Earthbanc employees suggest the legal dispute in Kenya is just the tip of the iceberg, with allegations that the startup inflated the project’s climate impact – with little net benefit to local communities.

Sources also say that a much larger sister project ongoing in India – managed by Earthbanc, funded by AstraZeneca, and with ties to a conservation foundation established by King Charles III of England – is plagued by many of the same issues.

From lab to land

In 2019, high above Lake Geneva at the Caux Palace Hotel in Switzerland, Australian green entrepreneurs Tom and Chau Duncan, together with environmental specialist Rishabh Khanna, launched Earthbanc.

Then called ‘Goodbanc’ – the trio pitched a turnkey solution for corporates looking for verifiable carbon credits. Its core offering was Measurement, Reporting, and Verification (MRV) tech that uses satellite and drone data, AI, and remote sensing to verify that a project sucks in as much carbon from the air as it claims to. The company would manage projects on the ground. In a later pitch to investors, Tom Duncan also promised investors “forward contracts,” which would allow them to pre-purchase carbon credits, often at steep discounts.

The Caux Palace in Switzerland where Earthbanc was first launched in 2019. Credit: Wikimedia/Creative Commons

The pitch worked. Earthbanc secured a partnership in 2022 with the United Nations Convention to Combat Desertification (UNCCD) to launch carbon pre-purchase agreements to finance carbon credits. “This is part of climate justice,” Tom Duncan told TIME magazine shortly after winning the deal. “As carbon prices increase, people most vulnerable to climate change deserve that compensation to enable them to adapt to a changing climate and ensure they have food and water security.”

Investors followed. Later in 2022, the company closed a €1.5m seed round from the European Space Agency and venture capital firms including Rampersand, Katalista Ventures, Regenerative Ventures, Regen Network, Visive Capital, and Kaai Capital. And shortly after that, the startup landed its first major commercial deal with AstraZeneca. Contracts in India and Kenya made up the largest share of the company’s €17.7m revenue in 2024, a fivefold jump from €3.6m in 2023. Tech publication Sifted recently listed Earthbanc among Europe’s fastest-growing startups.

Backed by prominent investors and armed with a plan to remake a troubled carbon credit market, at the beginning of 2024 Earthbanc moved its technology from the lab to the field.

The fallout

In Kenya’s highlands, smallholder farmers typically earn a living growing maize, beans, and other crops on plots averaging a few acres. However, the launch of several carbon credit schemes in the region has opened up the promise of an attractive new source of income.

One key figure in the region’s bid to leverage carbon markets to bring revenues to local communities is Festus Kiplagat, GPI’s founder and a local to the region.

Kiplagat was first introduced to Earthbanc in 2022 through a mutual connection. He describes a productive initial partnership, which led to some half a million trees being planted in the summer of 2024.

A participant plants local green plants in a park as part of Ethiopia's Green Legacy Initiative. Tree-planting has become hugely popular across Africa in recent years. Credit: AP Photo/Amanuel Birhane/TT

​​Kiplagat says the relationship with Earthbanc began to unravel when he discovered that the company had quietly incorporated a subsidiary called Earthtree to take over the project's operations.

"You come to Kenya, I welcome you to my house, we agreed to do stuff together. And then unknown to me, you're building a different house right beside me," he tells Impact Loop in an interview.

The consequences, Kiplagat alleges, were far-reaching. Rather than sourcing seedlings from local community groups as was the case originally, Earthbanc started procuring them from its own centralised Earthtree nursery – effectively redirecting money meant to benefit local communities back to itself.

At the same time, he alleges, Earthbanc began pushing planting methods and species that served its carbon targets rather than the farmers. The company favoured eucalyptus and mulberry "monocultures" – chosen, Kiplagat says, to "maximise carbon units."

It was a pattern that, according to insiders, extended well beyond the farms.

'Big promises'

A former Earthbanc employee describes a high-pressure environment at the young company.

“They made big promises to return money to investors,” said the person, who wished to remain anonymous, citing confidentiality agreements. “And I felt they were under a lot of stress trying to plant as many trees as possible and reach the [target] numbers.”

On strict deadlines to deliver, the former employee alleges that Earthbanc’s wholly-owned subsidiary, EarthTree, resorted to direct seeding methods to beef up the numbers of trees it could effectively plant.

'They are fudging the data'

Direct seeding is a reforestation method where tree seeds are sown directly into the soil. It’s cheaper and faster than growing seedlings in a nursery and then transplanting them into the ground. However, it’s typically less effective. One study found less than 0.3% of seeds successfully germinated through direct seeding methods.

Recordings of internal meetings obtained by Impact Loop suggest a stark gap between Earthbanc’s ambitions and conditions on the ground. In the audio, CEO Tom Duncan expresses frustration over the project’s faltering progress, citing difficulties in implementing the technology as intended. He also laments feeling compelled to flout “every single contractual rule” governing how the trees are planted.

A spokesperson for Earthbanc counters that the project was monitored by “independent 3rd party forestry experts” and that “feedback was integrated in our workflow for continuous improvement.”

Trust breaks down

Kiplagat alleges the farmers in Uasin Gishu were initially presented with contracts that gave them legal ownership of their land but assigned the carbon rights to the project developer for 30 years, with an exit clause after 5 years.

In early 2025, Earthtree re-approached the farmers with new terms, according to a contractual agreement seen by Impact Loop – and offered a new 44-year contract without an explicit exit clause.

"We were going from farm to farm trying to convince landowners to sign these contracts so we could access their land for decades," another former Earthbanc employee told Impact Loop. "As you can imagine it wasn't easy."

With the new contract, and Earthtree in charge, trust between the company and local farmers quickly dissolved, Kiplagat claims.

Local farmers staged multiple small protests at Earthtree offices in the first half of 2025, according to local news reports.

Court case

In May 2025, Kiplagat sued Earthbanc and its subsidiary Earthtree for alleged breach of contract.

“The issue is that Earthbanc wanted to sideline [GPI] so that they benefit financially,” he says. “They established Earthtree to take over the project and cut off the community from benefits.”

A spokesperson from Earthbanc denies wrongdoing, arguing that GPI was “non-exclusive subcontractor” instead of an equal partner and that the company had paid the organisation “in full for services rendered.”

The spokesperson tells Impact Loop that while the company regrets that the specific project did not reach full scale, it remained “committed to applying these hard-earned lessons to our future restoration efforts globally.”

We first wrote about Earthbanc’s court case in July 2025. In response to that article, co-founder Chau Duncan said in an email that the company was “being targeted by criminals in Kenya who are trying to harm and defame the company in a shakedown, to try and extract funding out of the company, since we fired those contractors [GPI]. This is a common tactic in Kenya against foreign companies.”

Kiplagat refutes the claims. The case is now before the High Court in Eldoret, the capital of Uasin Gishu County.

India could be next

Three sources currently directly involved in Earthbanc’s flagship Regeneration Meghalaya project in India paint a picture that bears many similarities to the Kenya scheme.

“I know a polygon where they [EarthTree] did 7,000 direct seeding… only four survived,” one of the people says.

AstraZeneca has committed $71m to Regeneration Meghalaya, which aims to plant over 60 million trees by 2030 in the mountainous region of north-east India. The project falls under the Circular Bioeconomy Alliance’s “Living Labs” initiatives, a network of environmental programs launched by King Charles III in 2020.

'There are high stakes involved...powerful politicians'

Tree survival rates at Regeneration Meghalaya, including those planted from nurseries, were as low as 12% in some areas, the person claims.

“They [Earthbanc] are fudging the data,” another source claims. The person says they believe Earthbanc continues to benefit from the project even if tree survival rates are low, arguing it uses the scheme primarily to refine its technology – which includes machine-learning models that “learn” by being trained on real-world project data.

“If this thing [the project] fails, Earthbanc will still be okay, they will still have the technology...the IP,” the person says.

The three people allege that Earthtree has also bypassed local implementers – the Hill Farmer Shiitake Mushroom Cooperative and Worldview Impact – renegotiating landowner agreements to cut out the community-led organisations and, according to one of the people, moving to take control of the nursery operations that had been the local partners' core responsibility.

“There are high stakes involved…powerful politicians,” one of the people said when asked why AstraZeneca hadn’t yet closed down the Meghalaya project. “It would be a big embarrassment.”

The local price of carbon

In 2023, Kenya’s president William Ruto called carbon credits an “unparalleled economic gold mine” and his country's “next significant export.” Yet, while communities in East Africa hear promises of prosperity tied to carbon credits, they are often left grappling with complex systems, unclear agreements, and little to show for it.

“We tend to see a severe lack of transparency and benefit sharing in [carbon offset projects] involving local communities,” Lindsay Otis Nilles, a policy expert at Carbon Market Watch, tells Impact Loop. “There’s a lack of transparency around the project design documents, around the financial flows…because where’s the money going? Is it actually going to the communities?”

Last year, the Northern Kenya Rangelands Carbon Project, one of the world’s biggest providers of soil-based carbon credits to companies including Meta and Netflix, was suspended after a court found that it was established without proper consultation of Indigenous communities. Another, separate, soil carbon project in northern Kenya is also tied up in a legal battle.

Globally, more than 70% of carbon-offset projects were found to cause harm to Indigenous people and local communities, according to one analysis by Carbon Brief. Meanwhile, some 43% of the reports examined by the publication found evidence of carbon-offset projects overstating their ability to reduce emissions.

Kiplagat says he believes the “top-down nature” of some projects, including Regeneration Kenya, is a core part of the problem.

“If any carbon project does not benefit the local community, then it is dead on arrival,” he says.

Earthbanc’s spokesperson insisted that the Kenya project was based on “multiple consultations with landowners and communities,” including processes aimed at ensuring “Free, Prior and Informed Consent (FPIC),” a process which aims to ensure people are fully informed and freely agree before any project begins on their land or affects their rights.

For now, Kiplagat says he is in discussions with new partners to restart an agroforestry project in Uasin Gishu.

“There are good solutions out there,” he says.

If you have further information about this story, or something else you think we should cover, send us a tip at editorial@loop.se