We recently published a meta-assessment of 100 U.S. voluntary sustainability and impact reports from public and private companies, published over two years. None was filed under CSRD. We coded every report against the same set of fields using the same process. [Read the full assessment here.]
Previously, we looked at what the report said about Scope 3 disclosure. This article looks at assurance: who has it, what it covers, and what the latest trends say from our research.
Read more: Third-party assurance for sustainability reporting: What it is and why it matters
Of the 94 operating companies where we could determine it, 51 carry some form of external assurance.
Company size is generally the giveaway of who has it. Thirty-three of the 47 companies with more than $3 billion in revenue have assurance, compared with 10 of the 30 below that line. In voluntary reporting, that revenue gap matters more than whether a company is public or private, a finding we cover in more detail in the full assessment.
Of the 51 assured reports, 16 never name the assurer.
They describe the assurance instead: “an external third party has performed limited assurance,” “we obtain independent limited assurance of our scopes 1, 2 and 3 emissions annually,” “independent third-party limited assurance.” Often the assurance statement exists, but it’s on a corporate website, in a separate letter, or behind a link rather than in the report itself.
Large companies do this most. Of the 33 assured reports from companies with more than $3 billion in revenue, 12 don’t name the provider. Among assured companies below $3 billion, nine of 10 do name who provided the service.
One possible reason: at that scale, the assurance statement is a formal document with its own legal review, published separately. The effect on the reader is the same, though. Naming the provider adds credibility and trust to the report.
Of the 51 assured reports, 43 specify limited assurance, two describe a mixed engagement, and six never state a level. None carries reasonable assurance over all the data it publishes.
Limited assurance being the norm won’t surprise anyone in the field. It’s less expensive, faster, and appropriate for most sustainability data. The lack of any full reasonable assurance engagement in the sample is still notable as some regulations move toward requiring it. The six reports that claim assurance without naming a level leave out the detail that tells a reader how much checking was done.
Of the 51 reports carrying external assurance, one covers the whole document. That engagement is limited assurance, and it’s the only one in our sample conducted under AA1000 rather than ISAE 3000. AA1000 is designed to assess a report against a set of principles as well as verify its figures.
Twenty-five cover greenhouse gas figures only. Another 25 cover a defined list of selected metrics, typically emissions plus water, waste, or safety.
Technically, “externally assured” describes a narrow engagement. To a non-specialist reader, it can sound much broader. Someone who sees the phrase near the front of a report could reasonably conclude the whole report has been checked. In our sample, what was checked was almost always a carbon number and a short list of metrics. The narrative, goals, progress claims, and case studies carry no external opinion.
Assured reports are also more likely to include the structure that makes data verifiable. Forty-five of the 51 assured reports include a data appendix, compared with 21 of the 43 unassured reports. Forty-six include a framework index, compared with 26.
The pattern holds within both size groups. Among companies with more than $3 billion in revenue, 30 of 33 assured reports carry a data appendix, compared with eight of 14 unassured reports. Below $3 billion, the figures are eight of 10 and nine of 20.
These findings come from a larger meta-assessment of all 100 reports, covering emissions, materiality, goals, frameworks, and report design. [Read the full assessment here.]
Methodology: We reviewed 100 U.S. voluntary sustainability and impact reports published between September 2024 and September 2026, each coded against the same set of fields. None is a CSRD filing. Figures cover the 96 operating companies in the sample; denominators vary where a field couldn’t be determined from the report. Where a company’s data sits in a companion file we didn’t retrieve, its disclosure may be understated.
John Davies is the Director of Content at thinkPARALLAX. With more than 15 years of experience in journalism, editorial leadership, and sustainability communications, he specializes in sustainability reporting, thought leadership, and executive messaging.