Executive questionsWhat the institution needs to be able to answer.
- 01Which AI systems and activities could be subject to obligations under the Act?
- 02Who owns the applicability decision and how is that decision preserved?
- 03Which controls, reviews and documentation should be connected to each material AI system?
- 04What evidence would an authorised reviewer need to reconstruct the institution’s position later?
Operating modelTurn the requirement into institutional workflow.
01Inventory first
Regulatory analysis is unreliable if the institution cannot identify the AI systems, providers, use cases and owners that may be affected.
02Applicability is a decision
Cybatar can structure applicability analysis, but formal applicability should remain an authorised legal or compliance determination.
03Controls need evidence
Policies, reviews and technical controls should be connected to evidence requirements rather than maintained as disconnected documents.
04Version the position
Regulatory interpretations, system facts and controls change. Preserve the version, date, decision owner and evidence behind each material conclusion.
EvidenceWhat should be preserved.
- AI system inventory and ownership
- Documented applicability decision
- Control mapping and accountable owner
- Testing and assurance evidence
- Human oversight evidence where applicable
- Supplier and model documentation
- Decision and change history
Cybatar operating layerWhere the platform supports the operating model.
AI Estate and AI PassportsRegulatory Control GraphAuthorisation WorkflowEvidence FabricAudit RoomSupplier Assurance
This page is an operational research summary, not legal advice. Applicability and legal interpretation should be determined by authorised legal or compliance professionals.
Primary sourceEUR-Lex — Regulation (EU) 2024/1689 ↗Use the authoritative source for legal, standards or framework text. Cybatar provides an operating interpretation for institutional implementation.