A training record alone does not prove system-specific AI literacy. Article 4 requires that people deploying AI systems have the literacy needed for their specific role and system context.
EAB connects literacy evidence to each system, each role, and each obligation — making literacy part of the governance record, not a separate HR process.
Generic training records and policy acknowledgements do not satisfy Article 4. EAB makes literacy evidence structural — connected to the system and the role, not filed separately.
Every AI system has a literacy evidence section. Evidence is not filed in a generic training system — it is attached to the specific system record where the obligation arises.
AI System Owners need technical literacy. Business Operators need deployment context literacy. Supervisors need oversight-specific understanding. EAB differentiates by role — not a single checkbox.
Literacy evidence is not an HR process that runs parallel to compliance. It is a governance artefact inside the system record — visible in the obligation status and audit trail.
Literacy evidence completeness is tracked per person, per system, and per role. Gaps surface before audit. Missing evidence is an obligation gap — not a hidden training task.
When a system changes scope, purpose, or risk classification, literacy requirements may change. EAB surfaces literacy gaps when system changes trigger re-assessment.
At audit time, the literacy evidence record per system is exportable as a structured artefact — not assembled manually from HR files. The record shows what existed at the time of system operation.
Literacy evidence is not a separate workstream. It flows from system registration through to evidence readiness and supervisor approval.
When a system is registered and ownership is assigned, the platform generates the literacy evidence requirements per role. Requirements are based on the system's risk profile and deployment context — not a generic template.
Role holders upload their literacy evidence — training certificates, competency records, or structured declarations — directly in the system governance record. Evidence is timestamped and attributed on upload.
Supervisor approval includes a review of literacy evidence completeness. Approval without adequate literacy evidence is flagged. The final approval record shows which evidence was present at the time of sign-off.
System changes can trigger literacy re-assessment. Evidence records are maintained over time — the governance record shows what literacy evidence existed at every point in the system lifecycle.
A declaration is not evidence. Article 4 requires that AI literacy be demonstrable. EAB makes it structural.
EU-hosted · Anchored to CELEX 32024R1689
Tell us about your organization and what you’re looking to address. We’ll follow up with the relevant information.
