This essay is part of The Blueprint, a series exploring what contribution looks like in practice. It builds on The Default is Contribution, which argued that moral responsibility is the foundation (Part 1), contribution is architecture (Part 2), and the purpose of productivity gains is human flourishing (Part 3).
In the early 1990s, ornithologists at Keoladeo National Park in Rajasthan noticed that the vultures were disappearing. Not gradually. Catastrophically. Breeding colonies that had darkened the sky were thinning to nothing.
By the time the cause was identified — a 2004 paper in Nature by Oaks et al. — it was already too late to prevent one of the fastest population collapses of any bird species in recorded history. The culprit was diclofenac, a cheap anti-inflammatory drug given to cattle across South Asia. Vultures that fed on treated carcasses died of renal failure within days. The lethal dose was extraordinarily small. And diclofenac residues were present in roughly one in ten livestock carcasses — more than enough to destroy a population that reproduced slowly, one egg per year.
The white-rumped vulture, once estimated at 40 million individuals — possibly the most abundant large raptor on Earth — declined by 99.9 percent. Two other Gyps species fell by over 97 percent. India banned veterinary diclofenac in 2006. Two decades later, populations remain more than 95 percent below pre-crash levels.
Here is the part that matters for this essay: those vultures had been processing an estimated 10 to 12 million tonnes of carrion per year. They did it for free. No line item in any government budget. No entry on any balance sheet. The service was invisible — until it was gone.
Without vultures, carcasses rotted in the open. Feral dog populations surged by an estimated 5.5 million. Dog bites increased by tens of millions. Rabies deaths — already the highest in the world — rose by an estimated 47,000 additional fatalities over the crash period. Markandya et al., writing in Ecological Economics in 2008, estimated the total human health cost at roughly $34 billion. A veterinary drug costing pennies per dose had destroyed an ecosystem service worth billions per year.
No one had put vultures on the ledger. They appeared there anyway — not as an asset, but as a loss.
This is the pattern. The things that sustain everything we build are nowhere on the books. A forest that filters a watershed, a wetland that buffers a coastline against storm surge, a grassland whose root systems have been sequestering carbon since before the Roman Empire — none of these appear as assets. They show up only when they fail. When the watershed goes dry, the insurance claim arrives. When the coastline floods, the write-down follows. We have built a global economy on infrastructure we have never bothered to account for, and we are surprised that it is degrading.
Between 1992 and 2014, produced capital per person doubled and human capital per person increased by about 13 percent — but natural capital per person declined by nearly 40 percent. That finding, from the Dasgupta Review commissioned by HM Treasury, is the kind of number that should stop a CFO cold. It means we have been systematically liquidating an irreplaceable asset class, recording the proceeds as income, and calling it growth. Any fund manager would recognize this as reckless portfolio management. Applied to a company’s balance sheet, it would be fraud.
The World Economic Forum estimates that $44 trillion of global GDP — more than half the world’s economic output — is moderately or highly dependent on nature. Agriculture depends on pollination and soil biology. Pharmaceuticals depend on genetic diversity. Water utilities depend on intact catchments. Construction depends on timber and mineral cycles that function only within living ecosystems. And yet the financial system treats all of this as a free externality, an infinite subsidy from a planet that never invoices. Dasgupta goes further: the $44 trillion figure understates the problem because it measures dependency on nature’s flows — ecosystem services — while the stock generating those flows is itself shrinking.
Systems that consume capital without replacing it eventually collapse. The question is not whether nature will eventually appear on the ledger. The question is whether it will appear there by design or by disaster.
The Localization Trap
The harder problem — the one that makes biodiversity finance genuinely difficult in ways that carbon finance is not — is what I call the localization trap.
A tonne of CO2 is a tonne of CO2. It does not matter whether you avoided it in Manitoba or Malaysia. Carbon’s global mixing means the atmosphere does not care about geography, which is what makes carbon markets tractable. You can aggregate, you can trade, you can build fungible instruments across jurisdictions. The accounting is hard, but the physics cooperates.
Biodiversity does not cooperate. A hectare of lowland tropical forest in Borneo is not equivalent to a hectare of temperate broadleaf forest in Wales. A population of endemic freshwater mussels in the Tennessee River basin is not exchangeable for marine biodiversity in the Maldives. Biodiversity impact is irreducibly local — tied to specific species, specific habitats, specific ecological relationships that took millions of years to assemble and cannot be recreated elsewhere at any price.
This creates a problem for financial disclosure systems, which are designed to aggregate information at the firm level. The Task Force on Nature-related Financial Disclosures — TNFD — has more than 500 organizations representing $6.5 trillion in market capitalization committed to aligned reporting. But most of them stumble at the very first step: they cannot locate their assets relative to ecosystems with sufficient precision to say anything meaningful about their dependencies and impacts. A mining company may know it has operations in “Southeast Asia.” It may not know which watershed those operations affect, which species communities overlap with its footprint, or whether the river it draws from is in a biodiversity hotspot or a heavily degraded landscape where marginal damage is low.
Meanwhile, the data that does exist is deeply unreliable. Multiple studies have found that 35 to 55 percent of GBIF’s biodiversity records — the world’s largest open repository of species occurrence data — contain significant quality issues. Duplicates, impossible coordinates, records missing dates, specimens assigned to the wrong hemisphere. The foundation of global biodiversity monitoring is riddled with errors that make it effectively unusable for financial-grade analysis without significant cleaning and validation.
The funding gap compounds everything. The Paulson Institute estimates that biodiversity conservation requires $722 to $967 billion per year by 2030 to reverse decline. Current flows are approximately $133 to $154 billion — roughly a fifth of what is needed. And set against both figures is a number that should stop everyone cold: between $4 and $6 trillion flows annually to subsidies that actively damage nature. We are, as the Dasgupta Review put it, paying ourselves to destroy our own assets. Thirty to forty times more capital flows toward destruction than toward protection. This is not a market failure in the technical sense. It is a market functioning exactly as designed, optimizing for financial returns while externalizing ecological costs onto the future.
This is the broken ledger. And fixing it requires not just more money, but new instruments, new data, and new infrastructure. This is where AI enters.
What the Physics Now Permits
Something has changed in the last five years that Paul Hawken and Amory Lovins could not fully anticipate when they wrote Natural Capitalism in 1999, though Lovins’s fourth principle — invest in natural capital — was always pointing here. The monitoring problem that made biodiversity finance intractable is becoming tractable. Not because the politics have shifted, but because the cost curves have collapsed.
Environmental DNA analysis — detecting species from fragments of genetic material left in water or soil samples — has transformed biodiversity surveying. NatureMetrics, a UK-based company, demonstrated in field trials that eDNA detected 60 percent more mammal species than traditional camera traps, using just 40 water samples. A survey that would have taken weeks and cost tens of thousands of dollars can now be completed in days at a fraction of the cost. The limiting factor shifts from data collection to data interpretation, and that is where AI compounds the advantage. NatureMetrics now operates in more than 110 countries, and in January 2025 it raised a $25 million Series B co-led by Just Climate — the fund established by Al Gore’s Generation Investment Management — alongside EDF Pulse Ventures and the Monaco ReOcean Fund. Serious climate capital is moving into biodiversity monitoring.
Planet Labs operates a constellation of satellites that images nearly the entire land surface of the Earth every day at approximately three-meter resolution. Deforestation events that would have been invisible to financial disclosure systems until the next annual report are now visible the morning after they happen — peer-reviewed studies using satellite data and neural networks have achieved 89 percent accuracy in automated detection against Brazil’s official PRODES monitoring program. Passive acoustic monitoring, once limited by the sheer volume of data a single field station generates, has become viable at scale now that AI can process continuous, multi-species audio from thousands of sites simultaneously. Ecosystems can, for the first time, speak in something close to real time. And computer vision has crossed its own threshold: iNaturalist’s AI identifies over 109,000 taxa at roughly 89 percent accuracy, Cornell’s Merlin app identifies birds from sound recordings at over 90 percent — consumer applications, running on phones, generating a volunteer sensor network of extraordinary scale.
The NatureTech Alliance — launched at Davos in January 2024 by Planet, NatureMetrics, ERM, and Salesforce — is embedding continuous satellite and eDNA monitoring into corporate sustainability workflows at scale. The monitoring that was once reserved for major infrastructure projects is becoming a standard operating capability for any company with significant land footprints.
The policy infrastructure is catching up. The UK’s Biodiversity Net Gain requirement, which became mandatory for most developments in 2024, has created the first statutory biodiversity pricing system of scale, with statutory credits priced at £42,000 per unit — deliberately set high to incentivize the private market. Verra launched a biodiversity credit framework under its Nature Framework in 2025, one of the first major registry-backed systems. The voluntary biodiversity credit market is still tiny — roughly $8 million per year compared to the voluntary carbon market’s peak of approximately $2 billion in 2021 — but the infrastructure is being built, and infrastructure compounds.
Lovins was right. He was just early. The Dasgupta Review gave the argument institutional weight: nature is not an externality to be priced at the margin but an asset to be managed in the portfolio. The market is finally catching up to what Natural Capitalism predicted twenty-five years ago: that treating nature as productive capital rather than a free input would unlock enormous value. The technology to do it at scale simply did not exist yet.
The Contribution Test
This is where the analysis must be honest, because the history of environmental finance contains cautionary tales.
The three questions I carry from the triptych: Is this contribution or extraction? Is the architecture designed for flourishing, or is safety bolted on? Who benefits — and who bears the cost?
Applied to the emerging biodiversity finance ecosystem, the answers are mixed in ways that matter.
Start with what is now visible. The partnership between Bloomberg and the Natural History Museum London has placed the Biodiversity Intactness Index — covering more than 58,000 species, over a million physical assets, and nearly 50,000 companies — inside the financial data infrastructure that fund managers and analysts use every day. For the first time, a portfolio manager can pull up a holding and see its nature footprint alongside its earnings per share.
But visibility is not the same as value. Visibility is descriptive — Bloomberg can show you the score. Value is prescriptive — it changes decisions, redirects capital, shapes incentives. The ledger that matters is not the one that reflects what happened. It is the one that determines what will happen next.
The measurement, reporting, and verification (MRV) problem is still an open wound. According to industry estimates, thirty to forty percent of project revenues in early biodiversity credit markets are consumed by verification costs — the expensive fieldwork and third-party audits required to substantiate claims about ecological outcomes. This is an enormous tax on contribution. It means that a landowner in a biodiversity corridor who wants to generate credits and fund conservation is spending nearly half of every dollar they earn on proving to intermediaries that they did what they said they did. This is not a designed architecture for flourishing. It is exactly the pattern the triptych identified: safety — in this case, credibility — bolted on after the fact, at crushing cost.
The deeper risk is what I would call contribution-washing: making nature investable without making it flourish. The danger is not that the market fails. The danger is that it succeeds on its own terms — generating returns, satisfying disclosure requirements, flowing capital — while the ecosystems it purports to protect continue to degrade because the metrics selected are legible to financial systems rather than ecologically meaningful. A company can offset its biodiversity footprint by funding tree planting on degraded agricultural land while its supply chain continues to drive deforestation of primary forest. The credits balance. The ecology does not.
And on the question of who benefits: the answer must include the communities and ecosystems that have been subsidizing the global economy without compensation. Contribution-first biodiversity finance is not charity extended from capital markets to nature. It is overdue payment on a long-running debt. The indigenous communities who have stewarded biodiverse landscapes for generations while the global economy extracted value from them deserve to be on the right side of the ledger. Any architecture that makes nature investable while marginalizing those communities has failed the contribution test, regardless of its ecological metrics. India’s vultures fed a subcontinent for free. The communities that coexisted with them bore the cost when they vanished. The ledger was always being kept — just not by the people who profited.
The Blueprint
What does a contribution-first architecture for nature finance look like? Five components.
Standardize the foundation. The biodiversity data problem is solvable but requires coordinated investment. GBIF’s data quality issues are not a secret; fixing them requires dedicated funding for taxonomy, georeferencing, and duplicate resolution at scale. The community of practice around eDNA, acoustics, and remote sensing needs to converge on shared protocols so that data generated by different methodologies can be combined and compared. This is unglamorous infrastructure work. It is also the prerequisite for everything else.
Embed locality. Financial instruments for biodiversity must be designed around the localization trap, not in defiance of it. Jurisdiction-specific credit systems. Habitat-type classifications that resist inappropriate trading. Additionality requirements that account for baseline ecological condition. Resist the temptation to build a global biodiversity credit market on the model of carbon — the physics does not permit it, and attempting it will produce a market that is financially efficient and ecologically meaningless.
Compress verification costs. The MRV tax on contribution-first projects is the single most important problem to solve for market scale. AI-powered continuous monitoring — satellite, acoustic, eDNA at reduced cost — should be the primary verification methodology, with expensive human fieldwork reserved for calibration and dispute resolution. Drive verification costs below ten percent of project revenues and you transform the economics of conservation finance.
Price the externality at the margin. The $4 to $6 trillion flowing to nature-damaging subsidies annually will not stop because we ask nicely. It will stop when biodiversity damage is priced into the cost of doing business. Mandatory TNFD disclosure is a starting point. Statutory BNG requirements are a further step. The destination is a system where every significant financial decision carries a biodiversity price signal — not a bolt-on offset, but a real cost that shapes the decision from the inside.
Design for the communities. The architecture of nature finance should be designed from the outset so that the people most directly connected to biodiverse landscapes are primary beneficiaries, not afterthoughts. Payment mechanisms that reach smallholders and indigenous custodians directly, not only after institutional intermediaries have taken their margins. Governance structures that give those communities voice in how their ecosystems are valued and traded. Contribution-first means the contribution flows back to the source.
The ledger has always been kept. It was just kept by the future, in the currency of ecological degradation — biodiversity loss, watershed failure, climate instability, the slow unwinding of the systems that make complex civilization possible. Between 1992 and 2014 we doubled our produced capital while destroying 40 percent of our natural capital per person. We called that progress.
What AI and the technology revolution in nature monitoring have given us is the chance to keep the ledger differently. To make the invisible visible. To price the externality before it prices us. To design a financial system that finally reflects what Dasgupta formalized, Hawken understood, and Lovins predicted: that investing in natural capital is not a sacrifice of economic returns but the deepest source of them.
The companies building this infrastructure — the NatureMetrics and Planet Labs of the world — are not doing charity. They are building ledgers that tell the truth. They are, in the most precise sense, contribution-first: creating the conditions under which value can be returned to the systems that generate it.
The question for every business leader reading this is not whether nature will eventually appear on your balance sheet. It will. The question is whether you will be in the architecture meeting when the ledger is designed, or whether you will receive it as a regulatory requirement after the fact.
The only thing left to decide is which side of the ledger you want to be on.