Let me warn you in advance about the high level of sarcasm and acidity in today's newsletter. But I find it hard not to reach for those tools when the output of several hundred people's work, produced through exceptionally difficult compromises, work done to make reporting easier for companies, is butchered on the home straight so that someone can boast: "look how efficient we are — the moment the delegated regulation is out in English, we're ready with 23 versions in the languages of the Member States."
I hasten to inform you: dear Commission, you are not ready at all. What is more, rushing out unverified translations creates additional problems. Which ones? Here you go, three examples:
- ESRS S1, paragraph 20(c) requires, in the original, the number of "temporary employees" to be disclosed, while the Polish translation at that point refers to "pracownicy tymczasowi" (temporary agency workers). Temporary agency workers are an entirely different concept from employees on fixed-term contracts. The former belong to the category of non-employees, while the latter most certainly are employees. Applying the Polish translation of the standards, we would therefore be counting completely different persons than under the English original.
- Annex II contains a definition of "designed recyclability rate" ("wskaźnik zaprojektowanej zdolności do recyklingu"), which gives an example of an event at the use phase of a product that may reduce that recyclability. The original example refers to "contamination preventing recycling". The Polish translation, however, points to "recykling zapobiegający zanieczyszczeniu" — "recycling preventing contamination" — which reverses the causal chain. The point is surely not to design a product in such a way that recycling it will prevent some contamination. The point is that a product may become so contaminated during use that recycling it later will be impossible.
- ESRS 1, AR 14 to paragraph 32 sets out a list of information sources on which a company may base its materiality assessment. The original version includes "peer experience" among them. I think we all know who peers are; almost every company I know has carried out some form of peer review or peer analysis during its materiality assessment, i.e. a review of what is material for entities comparable to it (companies in the same sector or with a similar business model). The Polish version of the standards, meanwhile, speaks of "doświadczenie partnerów" — "the experience of partners". Which partners? Because in the ESRS the term "partner" is in fact reserved for two kinds of partners: business partners and social partners.
In the Polish translation of the simplified ESRS I identified a total of 30 errors that change the meaning of legal provisions, plus a further 118 terminological inconsistencies. The second group is somewhat less dangerous, but it will add work at the reporting stage. It occurs most often where a term invoked in the ESRS by reference to another piece of EU legislation was not taken over from the Polish language version of that act, but translated anew. The definition of "manufacturer of articles" in Annex II refers directly to the definition set out in the REACH Regulation. The Polish version of that regulation defines such a person as one who "wytwarza lub składa wyrób" — manufactures or składa (puts together) an article. The very same definition, as quoted in the ESRS, reads in Polish "wytwarza lub montuje wyrób" — manufactures or montuje (assembles/fits) an article. I do not want to prejudge whether the REACH translator chose the right word in "składanie". Perhaps "montowanie" should have been used back then. But I would certainly rather not spend hours, or even minutes, in discussions, first internally within a company and later with the statutory auditor, deliberating on the difference between "montowanie" and "składanie".
After the simplified ESRS were adopted on 3 July 2026, someone in the Commission's translation service simply dropped their text into some LLM, produced 23 language versions, and then published them. Without any verification or alignment with other legal provisions. The fact that the result is not as devastating as three years ago is, I suspect, due solely to progress in the development of artificial intelligence models. The procedure (assuming any quality control procedure applied at all) has not changed.
Throughout the 13 months in 2023 and 2024 during which we waited for the standards to be corrected, I raised this issue at every meeting with Commission representatives. Ad nauseam. Over the past year, first as we were approaching the final shape of the simplified ESRS at EFRAG and later while the Commission was working on them, I raised the issue again and called for the translation this time to be checked, verified and of good quality.
I submitted the list of 148 errors identified to the Commission last Friday. I have been assured that DG FISMA staff will do everything in their power to have the Commission's translation services introduce the appropriate corrections before the text goes for publication in the Official Journal. I hope that will happen; there is a chance, because the scrutiny period expires at the beginning of September or (if extended) November. I will certainly keep raising this topic, again ad nauseam and until it is resolved, because this particular antithesis of due diligence really does grate on me. The simplified ESRS are much easier to apply than their first version, so the Commission should not make companies' lives harder by publishing a poor-quality Polish translation 😊
Polish 
English