Questions we are actually asked

The EU importer. If you produce outside the EU, CBAM does not name you; it names the company bringing your goods in.

That is why it reaches you as a questionnaire rather than a regulatory letter. Your customer has to declare the embedded emissions of what they import and hold evidence for the figure, and they cannot do either without you.

For CBAM, yes — the scope is a fixed list in Annex I, and if your goods are not on it the mechanism does not attach to them.

It rarely means nothing changes. EU customers subject to CSRD have to report their own Scope 3, and the largest part of that is what they buy. The request arrives through procurement instead of customs.

We do not publish a figure for that, and you should be sceptical of anyone who does without seeing your production.

The gap depends on two things: the default the Commission sets for your goods, and your actual emissions intensity. Both are needed. What is generally true is that defaults are set conservatively so that missing data is never the cheaper option.

In the transitional period the obligation is to report. In the definitive regime the importer must also surrender certificates for the emissions declared.

Dates have been amended since the Regulation was adopted. Rather than quote one here that may age badly, we track what the EU actually publishes and link the official text — see the regulation radar.

Scope and timing have been amended repeatedly, including through the simplification package, so any summary written a year ago is unreliable — including one written here.

The dependable answer comes from the current consolidated text, which we link rather than paraphrase. What has not changed is the direction: more companies reporting more, with assurance attached.

Because it is the only part you cannot measure on your own site. It is typically the majority of the total, and the usual fallback — spend multiplied by an industry average — produces a number that barely moves when your actual suppliers change.

That makes it useless for reduction and awkward under assurance, because an auditor asks where the figure came from and the honest answer is "an average".

Auditors accept evidence, not systems. What we can do is make the evidence cheap to examine: every figure keeps its inputs, the emission factor with its source and year, and the calculation that connects them, so the auditor is checking a trail rather than reconstructing one.

We do not issue assurance and we are not an accredited body. Anyone telling you their software makes an audit unnecessary is describing something other than an audit.

Often they mean a product carbon footprint and use "LCA" loosely. Sometimes they mean a genuine multi-impact assessment, and occasionally they mean a verified EPD, which is a different exercise again.

It is worth asking which of those they need before producing any of them. The comparison page sets the four side by side for exactly this conversation.

Not today. The multi-impact engine is built, and carbon runs on it in production. What it lacks is a licensed inventory database covering water, acidification, eutrophication and resource use to the standard a real assessment needs.

We would rather say that plainly than produce a multi-impact number we cannot source.

There is no single date. ESPR gives the Commission the power to set requirements product group by product group through delegated acts, so the answer depends on when the act covering your products arrives. Textiles and batteries are early.

The work that takes time is not the passport format. It is being able to state, per product, what it is made of and where that came from.

It is a signed, machine-readable product record built on the structures the EU work points at — a W3C Verifiable Credential, a GS1 Digital Link, a public scan URL.

Whether a given passport satisfies a particular delegated act is a question about that act, and we will not claim conformity with a requirement that has not been written for your product group yet.

It is verifiable, which is not the same word. Every figure carries what is needed to re-run it — inputs, factors, method — and a way to confirm the record has not changed since it was made.

Verified means an accredited third party examined it and issued an opinion. Making a figure verifiable makes that examination faster and cheaper. It does not replace it, and we do not describe self-declared data as verified.

For one narrow job: publishing a hash somewhere we do not control, so that a record can later be shown to have existed unchanged. That is a property no database we own can offer about itself.

Nothing about your business goes on chain — only a fingerprint. The point is provenance, not the ledger, and if a better append-only timestamp existed we would use that instead.

Polygon, and the anchoring runs on the Amoy testnet. We say testnet because it is testnet.

Moving to mainnet is a deliberate step with cost and operational implications, taken with a customer rather than announced as though it had already happened.

Yes, and without asking us either. The verifier on this site runs entirely in their browser: they paste the certificate, the browser recomputes the Merkle proof and compares it with the anchored root.

They see the carbon figure and its provenance. They do not see quantities, prices, or anything else in the record.

On infrastructure in the country the data is generated in, isolated per company. It is not pooled with other customers and it does not leave that boundary — only tamper-evident hashes do.

The free tools on this site are a separate matter: they run entirely in your browser and send nothing anywhere, because there is no server behind them to send it to.

No. And the calculation path contains no model at all: a carbon figure is arithmetic over declared inputs and versioned factors, which is what makes it reproducible.

Where AI is used — reading documents, spotting anomalies, answering questions about regulation — it works around the number, never inside it.

All of them, on every plan. Multi-factor authentication, session hardening and rate limiting do not vary by what you pay; this is enforced in the code rather than promised in a table.

Plans differ in product capability — how much you can produce, whether records are anchored, API access. Never in whether your account is protected.

From published datasets, and every factor carries its source and year into the result. The free calculator on this site uses a 43-factor catalogue derived from the audited catalogue in the platform, with the version stamped on the output.

Some entries are weaker than others — an ageing polyester factor, some without a redistributable licence — and those are flagged rather than quietly averaged in.

That depends almost entirely on your data, not on the software. If your energy and production figures are already systematic, the first product footprints come quickly. If they live in spreadsheets across several people, that is the work.

The free calculator here is a reasonable way to find out which situation you are in before talking to anyone.

The tools on this site — the calculator, the CBAM check, the verifier — need no account at all.

The platform itself is operator-onboarded rather than self-serve. It holds production data under a tenancy model that we set up deliberately, not through a sign-up form.

We do not publish customer names or logos, and there are none on this site. When there are references to give, they will be real and approved by the company named.

In the meantime the honest substitute is the product: the calculator computes, the verifier verifies, and the regulation radar shows what it pulled and when.

RecoChainAI OÜ, registered in Estonia, with the engineering and first pilots close to the industry it serves in Türkiye.

It is early and founder-led, and the site tries not to sound larger than that.