The Carbon Credit Shell Game: Why Most Offsets Fail Integrity Tests

In 2021, Delta Air Lines spent $137 million on carbon offsets, promoted itself on flights, in press releases, and on in-flight napkins as "the world's first carbon-neutral airline," and for a period, the claim held. Then The Guardian published its investigation into Verra's rainforest credits, which found 94 percent likely phantom with no measurable climate benefit. Then the class action arrived. By March 2022, Delta had "fully transitioned its focus away from carbon offsets."

The voluntary carbon market peaked at approximately $1.9 billion in 2022 and has contracted sharply since. According to Ecosystem Marketplace's 2025 State of the Voluntary Carbon Market report, the market fell to $723 million in 2023 and then to $535 million in 2024, with transaction volumes down a further 25 percent year-on-year. Ecosystem Marketplace describes the current phase as a supply-side reboot driven by quality demands, with the drop reflecting structural repricing across the supply chain. The correction is happening because the foundational claim (that these credits represent additional, permanent emissions reductions) has repeatedly failed close examination.

Yet 43 percent of credit retirements in 2024 still came from low-rated projects, according to MSCI's 2025 analysis. A shrinking market where the low-quality segment remains the majority by retirement volume is pointing at something specific about where demand is concentrated and who is still transacting.

The Question the Market Was Built to Avoid

Additionality is the foundational test. A carbon credit is supposed to represent an emissions reduction that would not have happened without carbon credit revenue. When a project fails that test, the credit is worthless as a climate instrument. A buyer paying for it is paying for a claim, and the seller is paid to keep the claim plausible.

Research published in Science (West et al., 2023) examined forest conservation projects across the voluntary market and found that they systematically set inflated baselines, projecting deforestation rates that peer-reviewed regional data could not support. The error runs consistently in the direction of the developer's financial interest. Higher baseline threat means more "prevented" deforestation, more credits issued, and more revenue.

Bloomberg's investigation into Nature Conservancy projects found a starker version of the same failure. Several protected forests in the portfolio faced minimal deforestation pressure regardless of credit revenue. National parks, indigenous land rights, and economics that made logging financially unviable had already secured those forests. The credits were monetising the absence of a threat rather than preventing one.

Across different project types, the mechanism reveals itself clearly. Forest conservation projects in regions with falling deforestation rates claim credit for a trend that pre-dates their involvement. Renewable energy projects claim displaced fossil generation in markets where wind and solar were already cost-competitive without carbon revenue. The ICVCM drove home this point in August 2024 when it removed eight legacy renewable energy methodologies from qualifying for its high-integrity CCP label, covering approximately 236 million unretired credits and 32 percent of the voluntary carbon market. Legal compliance projects bundle existing regulatory requirements with the offset, charging a climate premium for something companies were already obligated to deliver.

The Incentive Architecture Behind the Failure

Project developers set their own baselines, hire their own verifiers, and sell into registries that earn fees from volume. Verifiers, paid by the party they assess, carry limited financial incentive to challenge aggressive assumptions. Registries earn more when more credits are issued. The entity with the most at stake in accurate measurement, the global atmosphere, has no seat in any of those negotiations.

This is why the Guardian's Verra investigation did not expose a small number of bad actors operating around a sound system. The incentives converge reliably on quantity over quality, on claims that satisfy the appearance of rigor rather than the substance of it.

CarbonPlan's analysis of California forest offsets found that wildfire risk had been persistently understated across projects, generating permanence claims worth billions in issued credits that the underlying forest buffers cannot cover. Permanence, meaning the assurance that stored carbon stays stored, requires verification across the project lifetime, and the initial modeling at most projects flatters the developer's position. The problem does not show up at issuance. It accumulates forward.

The ICVCM's methodology reforms in 2024 addressed the edges of this problem. The financial infrastructure that produces low-quality supply, where developers, verifiers, and registries share aligned interests in credit volume, has not fundamentally shifted after the reforms. Progress in standards is visible. The underlying incentive calculus has not moved with them.

What Corporate Buyers Are Now Facing

Delta, Disney, Shell, Gucci, Nestlé. These companies purchased millions of offset credits that external assessments found to deliver minimal or zero climate benefit. Their neutrality claims rested on those purchases, and the legal exposure has moved well past the theoretical.

The Delta class action (Berrin v. Delta Air Lines, filed May 2023) alleged that the airline's carbon-neutral branding constituted consumer fraud under California law. A federal court in the Central District of California allowed key claims to proceed, including those seeking injunctive relief, after the plaintiff alleged that Verra's certifications rested on fraudulent carbon reduction projections and that Delta either knew or should have known those certifications were inaccurate. Delta contested class certification as of early 2025, and the case remained active. Delta's eventual pivot away from offset-based claims entirely functions as a case study in the limits of the model.

Parallel actions are underway across Europe. ClientEarth and partner organisations wrote to 71 airlines, warning that offset-backed claims about sustainable aviation are likely to be unlawful under emerging frameworks. The EU's Empowering Consumers for the Green Transition Directive, adopted in March 2024 and enforceable from September 27, 2026, explicitly prohibits climate-neutral claims based on offset purchases rather than actual emissions reductions. The Commission withdrew the proposed EU Green Claims Directive in June 2025 after political support collapsed, yet the ECGT is already binding and addresses the core issue with greater precision.

On the US side, the SEC adopted its March 2024 climate disclosure rules, challengers immediately contested them in court, and the Commission stayed enforcement in April 2024. The Commission then formally voted on March 27, 2025, to end its defense of those rules, leaving them effectively dormant. California's SB 253, with CARB regulations approved in February 2026, carries a firm Scope 1 and Scope 2 reporting deadline of August 10, 2026. SB 261 faces a Ninth Circuit injunction from November 2025, with enforcement paused pending the appeal's resolution. For companies with California operations at the scale SB 253 covers, the compliance window advances on its own track, regardless of any federal outcome.

The risk categories for corporate buyers run across four compounding dimensions. Securities fraud liability follows where climate commitments were made to investors on the strength of credits that later failed scrutiny. Consumer protection violations attach to misleading environmental marketing claims. When investigative journalism reveals poor credit quality, reputational damage arrives faster than any communications response can contain. Purchased credits that lose legitimacy in regulatory reviews create stranded asset risk, and all four compound in proportion to the prominence the neutrality claim holds in external communications.

What Carbon Offset Integrity Looks Like in Practice

Offset verification quality varies considerably, and experienced buyers can distinguish the difference before purchase. That distinction has also become measurable in price.

In 2024, high-quality credits rated A or higher averaged $14.80 per tonne, against $3.50 per tonne for low-quality equivalents. A credit trading at $3.50 is priced that way for reasons the developer has no incentive to advertise. Low prices in the carbon credit market are almost never a discovery. It signals quality issues that closer analysis routinely confirms. By 2025, the spread between MSCI's quality-split indexes had widened further, with the premium for higher-rated credits growing more than 20 percent year-on-year.

CarbonPlan's project database rates credits on additionality, permanence, and verification quality. Projects scoring above 80 deliver measurably higher integrity than the market average and are the most defensible basis for corporate claims. Gold Standard for the Global Goals operates the highest rejection rate and the most demanding additionality assessment among major registries.

The Oxford Principles for Net Zero Aligned Carbon Offsetting provide the most authoritative guidance for corporate buyers, prioritising absolute emissions reductions first, preferring removal credits over avoidance credits, and requiring long-term storage with demonstrably low reversal risk. The Science Based Targets initiative's Net-Zero Standard takes the strictest position on sequencing, requiring 90 to 95 percent absolute emissions reduction before any remaining gap is addressed through offsets or removals. The standard treats offsets as a residual instrument, which is precisely the ordering most corporate strategies have inverted.

The same documentation failures appear across different project types. Diligent buyers flag verification reports with confidential methodology and no public disclosure; REDD+ projects claiming threat rates above peer-reviewed regional averages from sources like Global Forest Watch; renewable projects in markets where the technology was already cost-competitive without carbon revenue; and credits priced well below the $14 to $50 per tonne range that high-quality credits command.

The voluntary carbon market ran on a story, and the story was more attractive than the verification required to support it. Delta paid $137 million for that story in 2021 and spent the following years explaining why it no longer applied. The buyers still transacting in low-rated credits account for 43 percent of the 2024 retirement volume. They are buying the same story at a lower price, in a smaller market, with better-equipped regulators and sharper plaintiff attorneys than anything their predecessors faced.

The credit is only as good as the process behind it. Most buyers have not built that process, and the regulatory frameworks now under construction were designed to find out.

Disclaimer

A note before the FAQs: this article is editorial and informational. The Industry Lens does not provide business, legal, financial, or technical advice of any kind. Carbon market regulation, legal exposure frameworks, and verification standards are evolving rapidly, with meaningful differences across jurisdictions and company types. Before making any decisions based on what you have read here, do your own research and consult qualified professionals, whether that means a climate accountant, carbon market specialist, or legal adviser, depending on the nature of your exposure. We are here to sharpen the thinking. The execution is yours.