How Annotated is made.
Annotated exists because the sources for understanding a regulation are scattered. The text is on EUR-Lex; the case law is on CURIA; the supervisory decisions are on each authority’s site; the drafting history sits across Commission, Council, and Parliament documents; the national implementing law is in the member-state legal databases; the official guidance is on the regulator’s website if you can find the right page. Each source is free. None of them are convenient. Practitioners do this work every day, and most of the time they do it from scratch.
Annotated does the assembly once, keeps it current, pin-cites everything, and hands you the regulation in context. That’s the work.
Sources
The annotated editions draw from:
- EUR-Lex for the canonical regulation text, recitals, corrigenda, and the official drafting-history documents (Commission proposal, Council positions, Parliament reports, final mandate-comparison).
- CURIA for Court of Justice and General Court judgments interpreting the regulation. We adopt CURIA’s own classification of outcomes; we don’t re-taxonomise.
- Rechtspraak for Dutch judgments that apply the regulation or its implementing law.
- Autoriteit Persoonsgegevens and the supervisory-authority registers across the EU for enforcement decisions.
- EDPB, the AI Office, and equivalent EU-level bodies for guidance.
- wetten.overheid.nl for the Dutch implementing law (UAVG and equivalents).
The bundling — and why it matters
The shift Annotated makes is from data to information to knowledge. The free sources give you the data. A search engine gives you the information. Knowledge is what happens when the right paragraph is next to the right court ruling is next to the right supervisory decision is next to the right guidance text is next to the right drafting note — pin-cited, named, dated, in context.
Tighter bundling beats broader coverage. Ease of access is the multiplier.
Taxonomy
We use the same classification the source uses. Court outcomes follow CURIA. Supervisory decisions follow each authority’s own categorisation. The data tags are not editorial inventions; that protects against the kind of opinion drift that makes legal databases unreliable two years later.
Update cadence
The editions are kept current as the law moves. Amendments and corrigenda get integrated as they’re published. Supervisory decisions and case law are ingested on a rolling cadence. Some sub-layers improve as we go — cite-resolution refinements, paragraph-grain backfill for older entries, edge cases in pin-cite shape. We name these as backlog where they exist rather than overclaim coverage.
Cross-references between editions
A reference from one annotated edition to another is a live link, not a citation. When you’re reading GDPR Article 6 and an entry cites AI Act Article 10, the link takes you to that paragraph of AI Act with the apparatus loaded. You cross publications without re-orienting.
What we don’t do
Annotated is a reference layer for practitioners doing the work. It is not legal advice. It is not a substitute for reading the source. It does not predict how a regulator will decide a case. The pin-cite tells you where; the conclusion is yours to reach.