Carbon-neutral labels meet the product’s emissions

A delivery can take the same route, in the same vehicle, burning the same fuel, and still acquire a different label. Buying carbon credits allows the seller to link that journey to a climate project somewhere else. Any additional climate benefit happens outside the delivery itself. The GHG Protocol’s product standard [8] preserves that distinction: purchased offsets are reported separately, rather than deducted from the product’s life-cycle emissions.

That boundary is now central to European consumer law. Directive 2024/825 [1] sets 27 September 2026 as the date for applying national measures that prohibit offset-based claims of a product’s neutral, reduced or positive climate impact. Implementation in France requires a qualification: the official records checked on 29 September do not establish that transposition has been completed. The EU deadline and the national legal position need to be kept distinct.

There is more at stake than the wording on a package. A credit connects emissions associated with a purchase to an intervention elsewhere, often running on a different timetable. Between the two sit a calculation method, verification and, sometimes, arrangements to replace a climate benefit if it is subsequently lost. The buyer receives evidence of a transaction. The climate outcome depends on the entire chain. Understanding the zero on the label means following the certificate back to the project and asking what actually changed.

The product still has the same history

Consider a hypothetical product responsible for 100 kilograms of carbon dioxide equivalent across its manufacture, transport, use and disposal. Carbon dioxide equivalent, or CO₂e, expresses the warming effects of different greenhouse gases on a common basis. All the numbers here are illustrative. The comparison assumes the same function, service delivered and life-cycle boundary.

In the first case, the manufacturer changes nothing. In the second, production is also unchanged, but the company buys credits associated with 100 kilograms of reductions or removals elsewhere. The product inventory still records 100 kilograms. In the third, a real change to production or sourcing lowers that inventory to 80 kilograms. Under these assumptions, the product’s emissions have fallen by 20%. Financing an external project could be added to that improvement, but would answer a different question.

Suppose the second manufacturer advertises zero after offsetting, while the third reports its 80 kilograms without trying to cancel them. A customer might reasonably infer that the second product causes fewer emissions. In this example, the industrial improvement belongs to the third. A useful intervention elsewhere and a change to the product itself become impossible to distinguish when a single number is expected to describe both.

Comparability matters just as much. A figure per kilogram of material, per garment and per year of use describes three different functions. The product standard [9] covers the full life cycle and explains why comparisons between products need additional specifications. Altering assumed service life or leaving out a stage can change the reported result without changing anything on the factory floor. The calculation behind the label deserves as much attention as the label itself.

From purchase to retirement

Several documents are casually called certificates. One may acknowledge the purchase of credits; another may record their retirement in a registry. A commercial certification may authorize the use of a logo. They do not necessarily establish the same thing.

Under Verra’s Verified Carbon Standard, one Verified Carbon Unit, or VCU, represents one tonne of CO₂e reduced or removed by a project under the programme’s rules. Issuance and retirement records are public. Verra’s description of these units [10] identifies what is being held: a registered unit tied to a project and a set of requirements, rather than ownership of a physically isolated tonne of gas.

The published process [11] requires activities to be implemented, outcomes monitored, a verification body to examine them and Verra to approve the verification before credits are issued. Units can then be sold. When used for offsetting, they are retired from circulation to prevent another use. Retirement refers to the credit’s status in the registry. It does not, by itself, describe CO₂ being extracted from the atmosphere.

The money can take different routes too. A purchase may pay the project developer, or a holder reselling units that have already been issued. A registry entry and a commercial contract therefore provide different information. To find out how much of a payment funds operations on the ground, one needs to identify the seller, the intermediaries and their commitments. There is no universal fee percentage that can be inferred from a certificate alone.

Paying before a project delivers raises a further question: what happens if the expected outcome never arrives? A forward purchase commitment and an issued credit can support the same project while giving the buyer quite different rights. Due diligence needs to establish what has been delivered, what remains outstanding and who must act if performance falls short. An attractive programme name cannot settle those contractual questions.

The credits depend on a world we cannot observe

For some projects, the decisive comparison is with an outcome that cannot be directly observed. A forest remains standing. How much would have been cleared without the intervention? An installation consumes less energy than the old equipment. Would the replacement have happened anyway?

This is the problem of additionality. The claimed benefit should depend on the financed intervention, rather than reward a change that would have occurred without credit revenue. The ICVCM’s Core Carbon Principles [12], a voluntary integrity framework, also require conservative quantification, arrangements for reversal risk and no double counting. A project can exist, operate and provide a useful service while leaving open the question of how much additional benefit should be credited to it.

Take another hypothetical example. During a given period, a project emits 40 tonnes of CO₂e. The calculation assumes that emissions without the project would have reached 100 tonnes. The difference is 60 tonnes. Change the assumed baseline to 60 tonnes and the calculated benefit falls to 20 tonnes. Observed project emissions remain at 40 tonnes in both cases.

The attributed benefit has fallen by two-thirds without any change to the observed outcome. The figure does not identify the right baseline. It shows why that assumption helps determine the volume that can be sold. A real assessment would also need to account for emissions displaced elsewhere and uncertainty in the underlying method.

For a forest project, one might need to ask whether logging prevented in one area moved to another. For a more efficient appliance, actual use matters, not just laboratory performance. The task is to establish a causal effect: what the financing changed relative to a credible alternative trajectory. A serial number can help prevent duplicate use of a unit. It cannot reconstruct the world without the project.

Competition can reward an optimistic baseline

Imagine two developers with identical total costs. One can justify twenty units; the other can justify sixty using a more favourable baseline. The second can offer a lower price per unit without becoming more efficient. A buyer sorting offers by price could end up choosing an accounting assumption while believing it had found a better climate intervention.

This example describes an incentive problem without alleging fraud by any company. Buyers may act in good faith, demand safeguards and finance sound projects. But a commercial interest in issuing larger volumes makes the quality of the underlying method crucial. Price competition works differently when the amount delivered partly depends on a scenario nobody can measure directly.

A review published in Nature Communications on 14 November 2024 [13] synthesized 14 studies covering 2,346 projects, supplemented by 51 studies of similar interventions. The authors estimated that fewer than 16% of credits issued to the investigated projects represented real reductions. Results varied substantially. This was neither a measure of intentional fraud nor a verdict on every credit sold in 2026. It demonstrates why compliance with a methodology still needs to be tested against independent evidence of its effects.

Standards bodies can challenge methods themselves. On 6 August 2024, the ICVCM [20] denied its quality label to eight renewable-energy methodologies because additionality was not sufficiently established. The decision concerned those methodologies, not the usefulness of renewable electricity. A power station can deliver valuable energy and have been financially viable without supplementary carbon revenue. Judging the installation and judging the credit are different exercises.

A fire can reopen the file

Credits also represent different interventions. Avoiding an emission changes a trajectory that would otherwise have occurred. Removing CO₂ from the atmosphere and storing it is another operation. Duration introduces a further distinction: storage may persist, or some of the stored carbon may later be released.

The Oxford Offsetting Principles, revised in February 2024 [15], prioritize emissions reductions and recommend shifting the offsetting of residual emissions towards durable removals. They also support protecting existing ecosystems for their climate and biodiversity benefits. A forest need not guarantee that it has cancelled an industrial emission to deserve funding. The difficulty arises when those two commitments are treated as interchangeable.

Reversal risk now has financial products of its own. On 25 August 2026, Verra approved an Artio policy for its durability pilot [18]. Launched in December 2025, the programme tests insurance and fund-based approaches as alternatives to a pooled buffer of units. Projects must be admitted to the pilot. The announcement does not mean that all Verra credits carry this insurance.

The pilot’s rules [17] explain what happens after a confirmed reversal. Affected units are marked as reversed until they are replaced. They cannot be transferred or retired while carrying that status. Without replacement, the marking remains. Recognizing a loss and restoring coverage in eligible units are therefore separate steps.

Consider a hypothetical forest loss. A replacement mechanism could preserve the crediting commitment by supplying eligible units from another project. Someone still has to mobilize the resources, obtain the units and satisfy the contractual conditions. If several projects suffer losses together, the availability of replacement capacity becomes important. This is a theoretical risk scenario; no loss or insurance shortfall in the pilot is established here.

Assessing the protection means examining its size, duration, exclusions and the replacement promised. A cash payment may compensate a financial loss without automatically delivering the same climate outcome. The certificate used in a marketing campaign may have been obtained long before these responsibilities are tested. This is where the contract meets the project’s longer life. Risk sharing between insurers and reinsurers requires a similar examination of commitments and their continuity.

The EU deadline and French law

Directive 2024/825 [1] addresses what consumers are told about a product. Its new point 4c targets claims that offsetting gives a product a neutral, reduced or positive greenhouse-gas impact. The practice is added to the annex of commercial practices considered unfair in all circumstances. The issue is therefore not simply whether a credit is expensive or well designed. It is what the seller claims to have changed about the product.

Recital 12 distinguishes this advertising from communication about environmental investments, which remains possible provided it is not misleading. There is no blanket ban on buying credits. Nor does a private label, by its existence alone, authorize a claim caught by the provision. Project certification and permission to use a consumer claim are separate tests.

France requires a careful qualification. The European Commission called for full transposition on 28 May 2026 [6]. As checked on 29 September, the National Assembly [4] and Senate [5] records still show a first reading in the Assembly following Senate adoption on 18 February. These records do not confirm promulgation completing transposition. A directive places obligations on member states; its application date should not be presented here as confirmation that transposition into French law is complete.

In the version of Environmental Code article L. 229-68 [3] examined, carbon-neutral advertising remains conditional on publicly available direct and indirect emissions information, quantified reduction plans and details of compensation. The French consumer authority, DGCCRF, describes the intended change [2]: moving from this disclosure-based regime to a specific prohibition for the claims concerned. Existing rules against misleading commercial practices remain relevant. This documentary analysis does not clear any particular campaign for use.

Older packaging needs care too. The common enforcement approach described by DGCCRF on 28 July [7] calls for proportionate assessment of genuine difficulties affecting products already on the market. It is not a general postponement of the reform. Finding an old logo on a shelf does not establish an unlimited exemption or demonstrate that every use of it is lawful.

The certification business is already adapting

There is a documented commercial response. On 22 April 2026, Climate Impact Partners [16] said clients using CarbonNeutral certification for products within the directive’s scope would have to move to its climate-contribution certification. It also introduced additional independent verification. The provider cautions that its guidance is not legal advice and that clients remain responsible for compliance. The supplier is adapting its offering based on its reading of the law; no regulator is certifying those claims here.

Changing the promise can matter. Saying that a company funds a particular project leaves the product’s emissions open for examination. Consumers can assess that contribution on its own merits instead of treating it as the disappearance of their purchase’s footprint.

Presentation will determine whether the distinction is visible. If the imagery, layout and headline still imply that a purchase has no climate consequences, a cautious line in small print may do little to explain it. Separately reporting the product inventory, the reduction trajectory and external funding would make different outcomes visible. A specific campaign must be examined as a whole; its legality and clarity cannot be inferred merely from a new label name.

The change could also alter commercial value. If a credit no longer supports a particular advertising claim, its usefulness to a buyer seeking that claim may change even while the underlying project continues to operate. The existence of the unit, its climate quality and its marketing value are distinct. This investigation has not measured a fall in demand or prices caused by the reform. It identifies a possible exposure to changes in demand.

Financing something that needs to last

That dependence on demand matters to developers as well. In a perspective published in PNAS Nexus on 12 May 2026 [14], Benedict Probst and Florian Egli argue that volatile revenue and uncertain demand make capital-intensive removal projects harder to finance. They propose more stable financing arrangements. Their article presents an argument and policy options, rather than demonstrated results from an implemented replacement system.

A hypothetical removal facility makes the timing problem clear. It must pay for equipment today and sell a climate service for years. Customers committing only for their next advertising campaign offer a poor match for that investment horizon. A long-term purchase agreement could reduce uncertainty, provided it specifies the deliverable and allocates responsibility for failure. Upfront funding can enable construction. Verification still has to establish what the facility actually delivers.

A similar distinction helps with ecosystem protection. Monitoring, community safeguards and alternative livelihoods can be valuable activities. The TFFF for tropical forests illustrates another financing architecture for conservation. Their purpose is easier to understand when it does not depend on generating exactly as many credits as a customer wants to deduct from its own footprint. Since 2021, France’s environment agency ADEME [19] has advocated distinguishing reductions within an actor’s own sphere from contributions to a collective goal, while recognizing the value of financing action elsewhere.

For a consumer, the decisive information comes down to concrete questions. Which emissions associated with the product have actually fallen on a comparable basis? What project received funding, for which period and with what verified outcome? Who must act if that benefit is lost? These answers can sit alongside each other on a clear product page. Compressing them into zero hides the differences needed to make a choice.

A certificate can document useful action. It cannot rewrite the life cycle of the product it accompanies. The project’s quality deserves examination; so does the product’s improvement. Keeping them visible separately makes it possible to follow the money without giving it credit for effects it has yet to establish.

Sources and documents

1. Parlement européen et Conseil / European Parliament and Council : Directive (UE) 2024/825

   2024-02-28. Accessed 29 September 2026.

2. DGCCRF : L’arsenal juridique de la lutte contre l’écoblanchiment bientôt complété

   Accessed 29 September 2026.

3. Légifrance : Code de l’environnement, article L. 229-68

   2021-08-25. Accessed 29 September 2026.

4. Assemblée nationale : Projet de loi DDADUE, dossier DLR5L17N53140

   Accessed 29 September 2026.

5. Sénat : Projet de loi DDADUE, dossier pjl25-118

   Accessed 29 September 2026.

6. Commission européenne, représentation en France : La Commission demande à la France de transposer la directive sur les règles donnant aux consommateurs les moyens d’agir en faveur de la transition écologique

   2026-05-28. Accessed 29 September 2026.

7. DGCCRF : Nouvelle directive sur la transition verte : l’approche commune des autorités pour contrôler les produits déjà sur le marché

   2026-07-28. Accessed 29 September 2026.

8. WRI / WBCSD, GHG Protocol : Product Life Cycle Accounting and Reporting Standard

   2011. Accessed 29 September 2026. Chapter 11, p. 87; Appendix A, p. 115.

9. GHG Protocol : Product Standard

   Accessed 29 September 2026.

10. Verra : Verified Carbon Units (VCUs)

    Accessed 29 September 2026. Operator source: attributed rules and announcements.

11. Verra : Verified Carbon Standard

    Accessed 29 September 2026. Operator source: attributed rules and announcements.

12. Integrity Council for the Voluntary Carbon Market : The Core Carbon Principles

    Accessed 29 September 2026.

13. Probst et al., Nature Communications : Systematic assessment of the achieved emission reductions of carbon crediting projects

    2024-11-14. Accessed 29 September 2026.

14. Benedict S. Probst et Florian Egli, PNAS Nexus : Fixing carbon credits requires a new financing model

    2026-05-12. Accessed 29 September 2026.

15. Université d’Oxford / University of Oxford : Oxford researchers launch updated carbon offsetting principles

    2024-02-28. Accessed 29 September 2026.

16. Climate Impact Partners / CarbonNeutral : What is the ECGT Directive and who is impacted?

    2026-04-22. Accessed 29 September 2026. Operator source: attributed rules and announcements.

17. Verra : Verra’s Approaches to Durability

    Accessed 29 September 2026. Operator source: attributed rules and announcements.

18. Verra : Verra Approves Artio Insurance Policy for Use under Durability Pilot

    2026-08-25. Accessed 29 September 2026. Operator source: attributed rules and announcements.

19. ADEME : Tous les acteurs doivent agir collectivement pour la neutralité carbone, mais aucun acteur ne devrait se revendiquer neutre en carbone

    2021-04-01. Accessed 29 September 2026.

20. ICVCM : Carbon credits from current renewable energy methodologies will not receive high-integrity CCP label

    2024-08-06. Accessed 29 September 2026.

Method and limitations

Documents checked on 29 September 2026. EU deadlines, French parliamentary records and national requirements are distinguished. This analysis does not clear any particular advertising campaign. Verra and Climate Impact Partners describe their own programmes; their statements are neither independent audits of all credits nor regulatory approval of marketing claims.

Both charts are hypothetical teaching examples, using no market data. The 2024 estimate concerns the projects studied, rather than every credit circulating in 2026. Possible effects on demand, prices and replacement capacity are risk mechanisms; no market collapse or pilot failure is established here. No interviews or direct exchanges with the actors were conducted for this article.