We built CarbTrust because the gap between what carbon credits claim and what they can demonstrate had become the single biggest obstacle to scaling durable removal.
By 2024, buyers had learned to discount every carbon claim by default — and they were right to. Investigations kept finding the same pattern: modeled numbers presented as measurements, verifiers selected and paid by the projects they audited, and raw data that nobody outside the transaction could see.
The casualty wasn't the bad projects. It was the good ones. A rigorously measured biochar operation and a paper-thin avoidance scheme were being valued by the same skeptical discount, because buyers had no way to tell them apart.
That's an infrastructure problem, not a morality problem. So we built the infrastructure: instruments we own and calibrate, verifiers assigned by rotation and paid by the registry, and every measurement published alongside the credit it produced. Not because we're more trustworthy — because the structure doesn't require you to trust us.
Most integrity failures in carbon markets aren't fraud. They're predictable outcomes of who pays whom. We changed the structure rather than adding another layer of promises.
These aren't values on a wall. Each one has cost us revenue at least once, which is the only real test of whether a principle is load-bearing.
If we can't put an instrument on it, we don't issue a credit for it. This rules out entire categories of project that other registries happily list — and it's the reason our inventory grows slower than it could.
Verifiers are rotated and registry-funded. Independence you have to take on faith isn't independence.
Free prior informed consent, documented benefit sharing, and community veto. We've rejected viable projects over this.
Avoidance credits are a different product with different math. Mixing them into removal portfolios is how claims get overstated.
Every methodology version is public and diffable. If we change how something is calculated, you can see exactly what changed.
Our founding team was hired to audit a corporate removal portfolio. Of 340,000 tonnes claimed, source-level measurement data existed for under a fifth. Nobody involved had done anything wrong — the data simply didn't exist to collect.
We stopped auditing and started measuring. Eleven biochar sites across Malaysia and Indonesia, fully instrumented, with raw data published from day one. The first verified batch issued that November.
We moved verifier assignment away from developer choice entirely — rotation from an accredited pool, funded by pooled issuance fees. Three developers left over it. The rest stayed, and pricing improved for all of them.
Biochar, enhanced weathering, direct air capture, and reforestation on a single measurement standard. 2.4 million tonnes verified, 180+ projects live, and every measurement still published.
They turned down our first application. The feedback was three pages long and completely correct. We fixed what they flagged, reapplied nine months later, and we're a better project for it.
Work with us
If you think the carbon market's problem is structural rather than moral, we'd like to hear from you.