What a passport field can and cannot prove

Can a DPP prove a product is sustainable, green or recyclable?

Can a Digital Product Passport prove a product is sustainable?

No, not by itself. A DPP can carry product information and the evidence or provenance around that information, but the existence of a passport does not turn the product into a substantiated sustainability claim. The conclusion still depends on what data exists, where it came from, what product and time period it covers, which method was used and which claim rules apply.

For example, a passport may contain a recycled-content value, a carbon figure or repair information where the applicable product rules call for it. That tells you what information is being asserted. It does not, on its own, establish the much broader conclusion that the whole product is sustainable.

Does a DPP prove a product is "green"?

No. "Green" is a broad environmental claim, not a status created by having a Digital Product Passport. A passport may make relevant product evidence easier to find, but the claim still has to be tested against its evidence, scope and the consumer-protection rules that apply to the words being used.

If the claim reaches further than the evidence, the passport does not cure the gap. A well-structured record can make that mismatch easier to see.

Can a DPP prove recyclability?

Not from a field alone. A DPP can carry information relevant to recycling, such as material composition, disassembly information or other end-of-life data where the applicable product rules require or permit it. Whether that supports a claim that this product is recyclable depends on the definition, method, product scope and evidence behind the conclusion.

The test is the same one used elsewhere on this page: identify the exact proposition, then ask what evidence would let another person reach the same conclusion. For the downstream information a recycler can actually use, see What a Recycler Can Actually Use.

What does a field in a DPP actually establish?

A field can establish that a defined value was recorded for a defined product object, ideally with its source, date and provenance. It establishes no broader conclusion automatically.

Use this sequence before turning a field into a public claim:

1. Data exists: what is the exact value or statement? 2. Source and evidence: who supplied, measured or calculated it, and what supports it? 3. Scope and method: which model, variant, batch or item does it cover, for which period and under which method? 4. Applicable claim rules: what must be true before those words can be used in the relevant market? 5. Conclusion: what is the narrowest statement the evidence actually supports?

So a DPP is better treated as an evidence-carrying product-information mechanism, not as an eco-label or a sustainability verdict. For the difference between product information and actual circular outcomes, see Digital Product Passports and Circular Supply Chains. For carbon and other modelled environmental values, use Carbon, Water and Environmental Data in a Textile DPP. Current environmental-claims status is maintained separately on EU Environmental Claims Rules from 27 September 2026 and What Actually Happened to the Green Claims Directive?.

Five tests between a value and a substantiation

These are not ranked by importance. They are ranked by how early they can be answered, because four of them can be answered from the document in your hand and the fifth needs somebody to compare the document with the sentence.

The last row is the one that decides cases. It is also the only one that cannot be answered by looking at the evidence alone, because it is a comparison between two objects: the exact words of the claim, and the exact scope of what you hold. Where two documents describe the same attribute and disagree with each other, that is a different problem with its own procedure, and it is set out on when the lab and the supplier disagree.

When the evidence had to exist

None of the five asks when you got the document, and there is a rule about that which most businesses discover late. Under the advertising code that governs marketing communications in the United Kingdom, the documentary evidence has to exist before the claim is distributed. It is a code rather than a statute, but as an operating rule it is the sharpest thing on this page. Evidence assembled after a challenge lands is not late evidence. On the code's own timing it is absent evidence, and the words went out unsupported whatever turns up afterwards.

Currency has a second failure nothing in your folder will show: the catalogue entry for the standard on self declared environmental claims records its edition as withdrawn on 24 June 2026 and replaced by a new one, so anything still citing the withdrawn edition now cites a document that does not exist (what a silent withdrawal does to a published field).

The standard the words have to meet

Substantiation has an operative standard. It is written as a standard about evidence and behaves as a standard about drafting. In the advertising code, an absolute environmental claim must be supported by a high level of substantiation. A comparative claim is treated differently: it can be justified where the product delivers a total environmental benefit over the thing it is compared with and the basis of the comparison is made clear.

Read that as an instruction about evidence and it sends a business off to buy more evidence. Read it as an instruction about drafting and it says something far more useful. The absolute form of a claim sets a bar the qualified form of the same claim does not, so the cheapest move available is nearly always to qualify the sentence rather than to fund the proof.

The register of decisions assembled for this estate holds fifty-nine records. Not one advertiser in it that set out to prove an absolute environmental claim is recorded as having succeeded. That describes the set we assembled rather than advertising everywhere, and the section below says how far it has been checked, but the direction it points in is consistent enough to plan around: absolutes are lost on their wording well before they are lost on their evidence.

The specification rule, which is the commercially important one

A claim counts as generic where its specification isn't given in clear and prominent terms on the same medium as the claim, so a link to a landing page doesn't rescue a generic claim on a label. Two cautions: the definition sits in the amended definitions article of the consumer empowerment directive, which we haven't yet checked word for word, so it's an open check rather than a settled reading; and clear and prominent is a standard, not a measurement, so nobody can tell you in advance how small the qualifying words may be. The rule in full: generic environmental claims from 27 September 2026.

In practice that means every claim on a product page traces back to the evidence behind it, and a claim with no traceable evidence does not get published.

The certification gateway has three limbs, not two

The list of practices treated as unlawful in themselves catches a sustainability label that is not based on a certification scheme or not established by public authorities, and a generic environmental claim where the trader can't demonstrate recognised excellent environmental performance relevant to the claim. That performance has three routes: the EU ecolabel, a national or regional type one scheme officially recognised in member states or top environmental performance in accordance with other applicable Union law. The third needs no certification scheme, so the choice isn't certification or silence. The third route rests on the amended definitions article, which we haven't yet checked word for word, so it's carried as an open publication check (how the estate handles a source it hasn't reached is on how we know); the three routes in full are on EU environmental claims from 27 September 2026.

A certification proves what it certifies and nothing adjacent

Where a scheme does stand behind a label, one sentence prevents most of what goes wrong next. A certification proves what it certifies. It proves nothing adjacent to what it certifies, however reasonable the adjacent thing sounds when somebody says it out loud in a meeting.

The failure cases all have the same shape. A certificate covering a fibre input carries a sentence about a finished garment. A certificate naming a mill carries a sentence about an order that mill did not make. A scheme that certifies a management system carries a sentence about a product. A certificate about restricted chemistry carries a sentence about environmental impact, which was never inside its scope. In each case the document is genuine, the scheme is real and the claim is unsupported anyway, because the sentence reached past the certificate's own subject. What each kind of document establishes before you ask it to support anything is set out on certificates, declarations and test reports. If the scheme itself changes, see what happens when a certification scheme is replaced.

Evidence fails on scope, and it fails on absence

A review of published decisions was assembled for this estate, and the pattern across it is consistent. Evidence was rejected because it addressed a different product, a different life cycle stage or a different question, or because there was none to produce. No regulator in that set was recorded as disputing the methodology of a life cycle assessment.

That is a statement about the set we assembled rather than about enforcement everywhere, and the next section says how far we have checked it. At that strength it still carries a consequence. Expensive, competent evidence loses when the sentence it is attached to reaches further than it does, and the money spent on producing it is spent either way.

So the highest return control in this subject costs nothing but discipline. Compare the exact words of the claim against the exact scope of the evidence you already hold, before anybody is asked to find more evidence. It is unglamorous, it is not a software feature, and on the evidence we have it would have prevented most of what we read.

A second, quieter failure mode sits underneath the first. A claim can be made where no evidence was ever produced, by a system that had no field for evidence and no way to record that the field was empty. A blank that carries no reason cannot tell anybody whether the evidence is missing, inapplicable or unrecorded. Putting a reason into that blank is among the cheapest improvements a passport project can make. To check one product, see which facts are usable, unsupported or missing.

A public claim can rest on restricted evidence

Both failure modes above assume the reader can see the evidence. Under a passport, often they cannot.

Four separate questions hide inside "is the claim substantiated":

The fourth is independent of the first three. Access control decides who can see evidence. It does not decide whether that evidence is sufficient, current, authentic or independently verified. A restricted evidence pack is not a stronger one, and a claim whose evidence nobody outside the company can see is not thereby better supported. It is the same claim with fewer people able to test it.

This changes who does the testing. Where evidence is public, scope errors get caught by readers. Where it is role-restricted, the only people who can catch them are the roles with access, and they may not be looking. The scope comparison this page recommends becomes more important as visibility narrows, not less.

Which roles see what, and how those permissions are meant to work, is a separate question with its own page. The battery access-right mechanics in particular are not settled: the implementing act that defines them has not been made, so any current description of who may see restricted battery evidence is provisional.

Why this page names no decision

The register behind the pattern above was assembled from mixed sources and it has not been closed. Twenty-six of its fifty-nine records sit unverified or only partly verified against the primary decision, and one claim drawn from that pool was refuted outright when somebody checked it independently.

A page that cited a decision it had not read, against the decision itself, would be doing exactly the thing it is warning about, one level up. So the examples are held back. When an item has been checked against its own primary decision and carries that decision's reference, it belongs here and it will appear with the reference attached. Until then the pattern stands and the illustrations do not.

Three specific attributions that were circulating in our own research record are not repeated on this estate in any form. An independent audit refuted the first against the public record, found the second internally implausible and could not source the third at all. They are not restated here, not as examples and not with the names removed, because an anonymised version of a refuted claim is still a refuted claim.

The mass balance question, stated and answered

Does a mass balance certificate let you say fifty per cent recycled on this particular item? No: it records a volume relationship across a supply chain, not the physical contents of the item somebody is holding, and a claim that omits which chain of custody model it rests on is quietly asserting the strongest of them. We haven't read the standard defining the models at its own text, and evidence arrives at batch or lot granularity while claims are made at item or model granularity, with no schema we've looked at carrying the rule between them (model, batch or item). What each model proves is on chain-of-custody models, and what a recycled-content value may claim is on recycled content.

Two different tests, enforced by two different bodies

Businesses selling in the European Union and the United Kingdom keep trying to write one sentence that satisfies both. There is no reason to expect that sentence to exist, because the two regimes ask different questions and are enforced by different bodies.

The European limb

A directive places four described practices onto a list of commercial practices treated as unlawful in themselves. Two of them are the provisions set out above: the sustainability label without a qualifying scheme behind it, and the generic environmental claim. Whether a claim is caught turns on the scheme, on the specification and on the wording, so this is a copy test rather than an evidence test.

What that list is, whom it reaches and when it attaches are on what applies today, which is where this estate keeps the applicability analysis. This page does not restate it and carries no date of its own for it.

The United Kingdom limb, which this page deliberately does not restate

The UK route runs through a different statute and a different enforcer. That is very nearly the whole of what this page says about the statute, because it reached us through a research summary rather than at its own text, and restating a legal test from a summary is the failure this page exists to warn about.

One position on the UK side does have a primary source behind it, because it comes from the regulator's own supply chain guidance rather than from the statute. It has two halves and it is only honest in both. Taking all reasonable precautions is not a defence, and an innocent breach is still a breach. And genuine compliance efforts may be treated as a mitigating factor when a penalty is calculated. Quoting the first half on its own makes a compliance programme look worthless, which is not what the guidance says. The same guidance also puts the sequence in an order many businesses do not: a retailer is expected to seek evidence ahead of advertising, and a brand is expected to provide assurance and to share the evidence underneath it.

Two things circulating that this page does not carry

Still a proposal, not withdrawn. The separate, later proposal on substantiating and communicating environmental claims is still formally open, awaiting the Council's first-reading position, though at least one commercial source publishes it as withdrawn. The check is a negative one at the primary source: the Official Journal notice listing Commission proposals withdrawn in October 2025 doesn't contain it, and the last formal procedural event recorded on the file was in December 2024. That check sits in our research record, not this page's source register, so it's named here rather than cited, and a file that hasn't been withdrawn isn't thereby progressing: is the Green Claims Directive withdrawn?

A judgment this page does not rely on. There is a judgment on the relationship between compliance with sectoral labelling requirements and the general regime on unfair commercial practices. Its text was never read at the court's own database because the primary route was blocked, so nothing written here rests on it in either direction. It is mentioned only so that a reader who has heard of it knows it is not the authority behind anything above. The reading of it that circulates most widely is the inverted one, as though it held that compliance with labelling law is no defence at all, and that is not a proposition this estate publishes.

What it looks like when somebody else fixed the scope

Most claims in this subject are written by the person making them, which is why scope drifts. Here is one where it does not.

The substances duty owed to a business receiving an article fixes every part of the claim in advance. The threshold is stated, the subject is the named substance, the minimum content of what is provided is stated, and the whole thing is capped at what the supplier actually holds. Nobody in that exchange gets to choose the scope, so the resulting statement is narrow, checkable and boring, which is what a well formed claim looks like.

The lesson transfers to claims nobody has drafted for you, and it is the same lesson the drafting standard and the specification rule both point at. Fix the scope before you write the sentence: which items, which market, which stage, which period. Then say what you fixed it to, in the same place you make the claim. A claim written scope first is harder to write and considerably harder to lose with. A certificate that supports a claim is also worth publishing where a machine can read it, which has exactly one available route today.

Sources

• Directive (EU) 2024/825 on empowering consumers for the green transition
  EU lawAnnex I points 2a and 4a; the definitions article as amendedCELEX 32024L0825In forceRelevant provisions reviewed

  The consumer empowerment directive. Two of the amended annex points were read at the Official Journal and the point numbering was independently probed. The amended definitions article, which carries the third limb of the certification gateway and the same medium test for a generic claim, has not yet been checked word for word. Both propositions resting on it say so in the sentence that uses them and both are recorded here as open checks. Nothing on this page states when the list becomes applicable, because the estate keeps applicability in one place.

• Digital Markets, Competition and Consumers Act 2024 (2024 c. 13)
  National lawIn forceReached through a secondary reproduction, primary text not read

  Declared so that the page can name the UK route without describing it. The statute reached us through a research summary rather than at its own text, so this page says that the UK test runs through a different instrument and a different enforcer, and stops there. Restating a test from a summary is the failure this page exists to warn about.

• Regulator guidance on supply chain evidence and environmental claims
  Regulator guidanceRelevant provisions reviewed

  The one thing on the UK side that this page states with a primary source behind it, and it is stated in both halves. Taking all reasonable precautions is not a defence and an innocent breach is still a breach; genuine compliance efforts may be treated as a mitigating factor when a penalty is calculated. The half quote makes a compliance programme look worthless, which is not what the guidance says, so both halves publish or neither does. The same guidance carries the ordering rule that a retailer is expected to seek evidence ahead of advertising.

• Advertising code provisions on holding documentary evidence before distribution
  Advertising codeRelevant provisions reviewed

  Carried for two provisions about substantiation rather than for the code as a whole: that an absolute environmental claim needs a high level of substantiation while a comparative claim is tested differently, and that the documentary evidence has to exist before the claim is distributed. It is a code rather than a statute, and the sentences that use it say so.

• Assembled register of advertising and consumer protection decisions
  Assembled register, partly verifiedRelevant provisions reviewed

  The basis of the pattern this page publishes and of nothing else. Twenty-six of its fifty-nine records sit unverified or only partly verified against the primary decision, and one claim drawn from that pool was refuted outright. So the set supports a description of itself at that strength. It supports no named decision, no party and no penalty. Three attributions that an independent audit refuted or could not source are not repeated in any form.

• ISO 14021, self-declared environmental claims, edition withdrawn 24 June 2026
  Withdrawn 24 June 2026Official source confirmed, detailed review pending

  Used for one fact and it is a fact about the catalogue rather than about any text: the edition on self declared environmental claims is recorded as withdrawn on 24 June 2026 and replaced. A technical pack citing the withdrawn edition now cites a document that does not exist. The replacing edition is behind a purchase, has not been bought and has not been read, so nothing here describes what it contains.

• REACH Regulation (EC) No 1907/2006, Articles 3(33) and 33
  EU lawArt. 33(1)CELEX 02006R1907-20250422In forceRelevant provisions reviewed

  The worked example of a claim whose scope is fixed by the instrument rather than by the writer. Read in full at the consolidated text. It is here because it shows what a claim looks like when somebody else has already decided the threshold, the subject and the minimum content.