CSRD Disclosure CheckerAmending Directive (EU) 2026/470, article by article

Method

How the record and the reading are made

The short answerThe record is read article by article from the saved EUR-Lex text of Directive (EU) 2026/470, each field cited to the article it came from. The reading matches a phrase specific to each requirement against your text, keeps your words verbatim, and says no mapping identified when nothing is reached.

The record

The amending directive was saved from EUR-Lex on 2026-10-17 as plain text and HTML. Each provision was read in the text, and its entry records the citation, the party it binds (read from the scope article), the requirement in our words, the evidence a regulator or an auditor asks for, the dates, the cross-references, and a source pointer on every field. Where the text states nothing, the record says not stated in the text. The record is not written from memory and not from a search snippet.

The reading

For each provision the record holds a short set of phrases specific to that requirement. The reading normalises whitespace, then looks for those phrases in your text and, where one is found, quotes the sentence of your text it came from. The status words are mentioned in your text and not mentioned in your text; they describe what the reading measured, not a compliance result. A sentence that reaches no provision is left unclassified and reads no mapping identified.

What the reading does not do

It does not decide whether the law applies to you, does not weigh materiality, does not audit and does not certify. Judgement stays with you, your counsel and your assurance provider.

Version and refresh

A reading names the date and the version of the text behind it. The text is re-downloaded and diffed against the saved copy, and the build fails loudly on any change, so the record cannot drift silently from the source.