For two years the argument about European sustainability reporting has been an argument about the future: how much it will cost, who will be exempted, whether the standards will survive simplification. Meanwhile a first group of companies simply did it. They filed sustainability statements under the new European standards, auditors gave opinions on them, and supervisors read them. That record now exists, and it is more useful than the forecasts. In May 2026 the European Securities and Markets Authority reported on 263 examinations of those statements carried out by national enforcers in 2025. In October 2025 it published a study of how 91 issuers explained their materiality assessments. EFRAG, the body that advises on the standards, gathered 656 assured statements. The French market and audit regulators each read their own selection.
This is the third piece of the sustainability chapter of the CareerOn Industry Atlas. The first asked where the money is and found that every euro figure in the public record is a forecast. The second asked who the chain is made of and found roles, not firms. This one asks the plainest question of the three: what did reporting companies actually do in 2025 and 2026, and what did the people who checked their work find? It matters to a candidate because the findings describe the work itself, the tasks a reporting team performs and the places where supervisors found it thin. We read eight dated documents from four families: European law, European and French supervisors, the French audit regulator and the standard setter. Every finding below is printed beside the sample it was read on, because a finding on a selection is not a finding on every company.
The first wave did the work and was read on samples: supervisors found the materiality step mostly done but often generic, the law delayed everyone except the companies already reporting, and no public document counts how many of them there are.
Eight documents, read for what was done
We admitted eight documents and refused four. The admitted set holds two adopted European acts from 2025, three ESMA documents, one EFRAG study, one study by the French market regulator and one inspection summary by the French audit regulator, the Haute autorité de l'audit. Four of the eight are new to this chapter. We refused a draft of the EFRAG study filed in board papers, because the final report exists. We refused copies of that report on other websites, because a copy is not the publisher's document. We refused a Council press release about the delay, because the act itself is admitted. And we refused an accounting recommendation that refers sustainability topics to ESMA without recording any finding.
Three rules governed the reading. A finding on a sample is printed with its sample. An adopted act is kept apart from a proposal, and nothing in this piece rests on a proposal. And where a number is a share of a collection, it is printed as a share, never turned into a count of companies. That last rule mattered more than we expected, as the section on who published will show.
Two laws of 2025, pulling in different directions
The law moved twice in 2025, and the two moves are easy to confuse. The first was Directive (EU) 2025/794 of 14 April 2025, known as stop-the-clock. It was aimed at the companies not yet reporting. In its own words, "the sustainability reporting requirements for those undertakings should be postponed by two years." The undertakings it means are those "currently required to report for financial years beginning on or after 1 January 2025 and on or after 1 January 2026." It gives its reason plainly: to avoid those companies "incurring unnecessary and avoidable costs" while the Commission worked on simplification.
The second was Delegated Regulation (EU) 2025/1416 of 11 July 2025, known as the quick fix. It was aimed at the opposite group: the companies already reporting. It did not stop them. It states that "undertakings that need to report sustainability information already with respect to financial year 2024 will be subject to a further phasing in of various reporting requirements with respect to financial years 2025 and 2026." In other words, the first wave keeps filing, with more requirements allowed to arrive later.
Two laws of 2025: one delayed the next waves, one lightened the first
| Act | Adopted | Who it reaches | What it does |
|---|---|---|---|
| Directive (EU) 2025/794, stop-the-clock | 14 April 2025 | Companies due to report for financial years 2025 and 2026 | Postpones their first reporting |
| Delegated Regulation (EU) 2025/1416, quick fix | 11 July 2025 | Companies already reporting for financial year 2024 | Phases in more requirements for 2025 and 2026; reporting continues |
Sources 1 EUR-Lex (Publications Office of the European Union) · 2 European Commission
The distinction matters for anyone reading headlines about a delay. The delay is real, but it did not reach the companies whose statements supervisors were already reading. Those companies reported for 2024, and under the quick fix they report again for 2025 and 2026. Neither act counts how many companies it covers. Stop-the-clock moves dates and names no one, and the quick fix lightens what is reported, not whether it is reported. The further change of scope proposed and adopted in 2026 is the subject of the first piece of this chapter and is not restated here.
Who read the first reports, and how many they read
Once statements were filed, five readers looked at them, and each read a different selection. ESMA's enforcement report counts the national work. "Of the 263 examinations of sustainability statements prepared in accordance with the CSRD/ESRS, 25% related to statements prepared voluntarily by issuers who were not yet subject to the ESRS because their Member States had not transposed the CSRD." A quarter of what enforcers examined, then, came from companies that reported before their own country required them to.
ESMA's materiality study read 91 issuers from 23 Member States, about one third of them applying the standards voluntarily. The French market regulator, the AMF, read 20 French listed issuers with more than 1,0007 employees, and said of its own work: "No statistical study (sample is not representative)". The French audit regulator studied certification work rather than reports. Its inspection summary describes « Les 16 mandats verts constituant le périmètre de l’étude », sixteen sustainability certification mandates held by ten audit units, selected across sectors rather than drawn at random.
What each reader of the first reports read, and on how many
| Reader | What it read | Sample | What the sample is |
|---|---|---|---|
| ESMA, with national enforcers | Examinations of sustainability statements in 2025 | 263 | Selected examinations; 25% were voluntary statements |
| ESMA fact-finding | Materiality disclosures in 2024 statements | 91 | Issuers from 23 Member States |
| EFRAG | Assured statements published to 20 April 2025 | 656 | Gathered without a complete public repository; France 16% |
| AMF | First statements of French listed companies | 20 | Stated as not representative |
| H2A | Sustainability certification mandates | 16 | Selected across sectors |
Sources 3 ESMA · 4 ESMA · 6 EFRAG · 7 Autorité des marchés financiers · 8 Haute autorité de l’audit (H2A)
The fifth reader is the standard setter, and its account of how it found the reports is the most revealing sentence in the record. "In the absence of a unique, comprehensive and public repository, the sustainability statements were gathered from a mix of data providers, alerts, advanced search, and other public sources." EFRAG, the body that advises the Commission on the standards themselves, had to search for the reports the standards produced. "This process returned 656 sustainability statements across various industrial sectors and geographies," published between 1 January and 20 April 2025.
Within that collection, "France (16%), Germany (13%), and Finland (12%) account for the highest number of CSRD-aligned disclosures." It is tempting to turn France's share into a number of French companies. We do not. Sixteen per cent of one collection gathered by searching is a share of what was found, not a count of who published. None of the eight documents we admitted counts how many companies in France, or in Europe, filed a statement for 2024. That absence is itself a finding: the regime can say what a sample did, and it cannot yet say how many complied.
Materiality: mostly done, often generic, examined twice
Every sustainability statement under the European standards rests on one judgement: which topics are material, either because the company affects people and the environment or because the topic affects the company's finances. The standards call this double materiality, and it decides what the rest of the report contains. It is also where supervisors looked hardest.
ESMA's study gives the clearest measure. "Around 60% of the sample met the overall objective of the IRO-1 disclosure requirement, which is to provide transparency on how the double materiality assessment was conducted." Read the other way, about four in ten of the 91 issuers did not explain well enough how they had decided what mattered. The enforcement report adds what that explanation looks like in practice: enforcers often found process disclosures boilerplate, with little sign of how a company had adapted the suggested steps to its own situation.
The enforcement report also measures a more mechanical part of the work. "A vast majority of issuers disclosed the list of ESRS disclosure requirements complied with (90%), and the list of datapoints derived from other EU legislation (85%), although, for the latter, only 70% reported complete information." That is preparation work, the lists a reporting team assembles before an auditor reads anything, and it was mostly done and sometimes incomplete.
Materiality was mostly done, often generic, and examined twice
| Finding | Share | Read in |
|---|---|---|
| Met the objective of explaining how the double materiality assessment was done | 60% | ESMA fact-finding, 91 issuers |
| Disclosed the list of disclosure requirements complied with | 90% | ESMA enforcement report |
| Disclosed the list of datapoints from other EU legislation | 85% | ESMA enforcement report |
| Reported that datapoint list completely | 70% | ESMA enforcement report |
| Materiality carried over as a priority for 2025 reports | Not a share | ESMA enforcement priorities |
Enforcement followed. The report records that "For 8 issuers, enforcers took enforcement actions in relation to materiality considerations in reporting under ESRS," and "For 4 issuers, enforcers took enforcement actions in relation to scope and structure of the sustainability statement." ESMA adds that enforcers acted "almost exclusively" by requiring a correction in the next sustainability statement, not a restatement, in a way that acknowledges the learning curve issuers are on.
Before the 2025 reports were even written, ESMA had already decided to look again. Its priorities for 2025 state: "ESMA has exceptionally carried over two of the priorities from its ECEP 2024, namely the implementation of the ESRS requirements on materiality and the scope and structure of the sustainability statement." Materiality was examined in the first year and announced as a priority for the second. The signal to reporting teams is continuity, not alarm: the same judgement, explained better.
The counter-case, kept in the piece
We set out expecting to show that materiality was the most frequent failing in the first reports. The evidence does not support that wording, and we have changed it. The enforcement report counts actions by priority, eight on materiality and four on scope and structure, and one action can cover several areas. No document we admitted ranks every area of finding against every other. Materiality is where supervisors chose to look twice, and it is the area with the largest count of actions among the priorities they set. It is not proven to be where companies failed most, and this piece does not say so.
The second limit is the samples. Every share above belongs to a selection: 263 examinations chosen by enforcers, 91 issuers chosen for a study, 20 French issuers described by their own reader as not representative, 16 mandates selected by an inspector. A selection chosen for risk may show more problems than the population, and a selection of large listed groups may show fewer. Either way, around 60% is a fact about 91 issuers, not about every company in Europe.
The third limit is absence. We found no public register that counts reporting companies, and the standard setter says it found none either. A national register may exist that we did not admit. Where we say no document counts publishers, we mean the documents we admitted, and we would print a count the day an official one is published.
What this means for your career
The first reports describe a job more precisely than any forecast could. Somebody in each reporting company had to run a double materiality assessment, explain how it was done in the company's own terms, and assemble the list of disclosure requirements and datapoints that the rest of the statement rests on. Supervisors examined exactly those tasks. The skill they asked for twice is not filling in a template. It is explaining a judgement: why these topics, on what evidence, with which stakeholders, at what threshold.
Two practice simulations on the CareerOn shelf now sit on evidenced steps of that work. ESG analyst: CSRD reporting and double materiality rehearses the assessment that ESMA's study and the enforcement report examined. ESG analyst: non-financial reporting under CSRD rehearses the preparation of the disclosure lists, the work the enforcement report measures at 90% and 85% disclosed, with only 70% complete. We place the second card on that preparation only. The record shows the lists were assembled and examined. It does not describe gap analysis as a named task, so we do not claim it does.
Two roles the record names have no simulation yet: the statutory auditor or independent assurance provider who gives the opinion, and the supervisory enforcement examiner who reads the filed statement. Both are recorded as gaps in the CareerOn catalogue. For a candidate choosing between them today, the evidence points one way. The first wave keeps reporting through 2025 and 2026, so the work these simulations rehearse is being done now, not after a delay.
For an employer, the reading is equally practical. The quick fix lightens what a first-wave company must disclose, not whether it discloses. A team that wrote a generic materiality section for 2024 is writing the next one under a supervisor that has announced it will look again.
What we could not prove
We could not prove how many companies published a statement for 2024, in France or in Europe, because no admitted document counts them. We could not prove that any share above holds for all reporting companies, because each belongs to a selected sample. We could not rank areas of finding, because no document does. We could not prove how many companies stop-the-clock covers, because the act moves dates and counts no one. And we could not say what the 2025 statements will show, because they are not yet examined. Each of these is a question we will answer from the record when the record answers it.
The CareerOn view, signed by the desk: the first wave proved the work can be done and showed where it is thin. A materiality judgement explained in the company's own terms is the skill supervisors asked for twice, and it is the one worth rehearsing first.