EU recitals as open data
Recitals explain why an article reads the way it does, and they are cited daily in questions of interpretation. Here they sit as an open surface: every recital has a number, a heading, an address, a hash and a reading date — enough to cite correctly and to notice when something changed.
What the surface gives out
- The act the recital belongs to, with its CELEX identifier.
- The recital number and heading, and the canonical address.
- Hash and reading date per row, so a copy can be checked against the original.
- A freshness score, so a reader sees how recently the row was re-read.
How to fetch it
- The address is /api/public/v1/open/eu/recitals and answers without a key.
- Responses carry ETag and Last-Modified; unchanged content returns 304 instead of the whole file.
- A next-check date travels with the response, so a fetcher knows when returning is worthwhile.
- Terms are the same as the rest of the open layer: use, cite, index, cache — do not resell the dataset as such.
The boundary
- The full body text of the recitals is not part of the open surface.
- Reasoning, bulk extraction and agent surfaces are priced as before.
- Metadata, citation string, hash and freshness are free and stay free.
What the usage shows
- NovaCopilot is used by professional lawyers and compliance teams in 17 countries.
- The agent API is used in product evaluations by international teams.
- AI Act articles are read in sequence — a clear sign of regulatory use.
Professional roles on this surface
Every row is its own node with source, reading date, hash and the same answer as data.
Workflows on this surface
Every row is its own node with source, reading date, hash and the same answer as data.
The articles the surface touches
Every row is its own node with source, reading date, hash and the same answer as data.
- GDPR — 5: Principles relating to processing of personal data
- GDPR — 7: Conditions for consent
- GDPR — 6: Lawfulness of processing
- GDPR — 13: Information to be provided where personal data are collected from the data subject
- GDPR — 22: Automated individual decision-making, including profiling
- GDPR — 32: Security of processing
- GDPR — 33: Notification of a personal data breach to the supervisory authority
- GDPR — 35: Data protection impact assessment
- GDPR — 46: Transfers subject to appropriate safeguards
- GDPR — 83: General conditions for imposing administrative fines
- AI Act — 6: Classification rules for high-risk AI systems
- AI Act — 5: Prohibited AI practices
- AI Act — 9: Risk management system
- AI Act — 17: Quality management system
- AI Act — 43: Conformity assessment
- AI Act — 49: Registration
- AI Act — 50: Transparency obligations for providers and deployers of certain AI systems
- AI Act — 71: EU database for high-risk AI systems listed in Annex III
- AI Act — 99: Penalties
- DORA — 17: ICT-related incident management process
- DORA — 28: General principles
FAQ
- May we use the surface commercially?
- Yes. Use, cite, index and cache freely. The only thing not permitted is reselling the dataset as such.
- How do we know a row is unchanged?
- Compare the hash. The same hash means the same row; a change also shows in the change journal with a date.
- How often are recitals updated?
- Every row carries its own reading date and freshness score, and the response states when the next check is reasonable.
Next step
Three ways to put the register to work in your own practice.
Start with your task
Litigation
Find support in a judgment
Search guiding decisions, see what became final and follow changes in the law.
In-house, deals
Map the rules in a transaction
Move from theme to act and on to the article that carries the duty.
Compliance
Assess the risk in a process
Risk scoring per legal area, with the sources behind every score.