EU AI Act Article 4: every training week produces a documented AI literacy record. Learn more →

Glossary

KI-Kompetenznachweis: Definition, Inhalt und Abgrenzung

KI-Kompetenznachweis erklärt: was in die Dokumentation gehört, warum ein Artefakt mehr belegt als ein IHK-Zertifikat oder eine Teilnahmeliste, Abgrenzung und EU-AI-Act-Bezug.

Short definition (citable, 45 words)

An AI literacy record is the traceable internal documentation of which AI literacy measure was delivered for which roles, systems and contexts of use, with what content and result. It is internal evidence that implementation happened. It is not an official certificate and does not on its own guarantee legal compliance.

Where the term comes from and how it shifted

The idea of a record is old in corporate learning. For safety or data protection briefings, companies have long kept records of who was trained on what and when. What is new is the concrete trigger. Article 4 of the AI Act created a literacy duty that must be demonstrable if questioned. That shifts the focus from certificate to record. A certificate is a seal, often attesting that a provider's exam was passed. A literacy record is the company's own complete account that an adequate measure took place for the right people. The law requires no seal but traceable adequacy. So in practice the AI literacy record replaces certificate-thinking. What matters is not a stamp but a solid trail.

The mechanism: what makes a record hold up

A record is worth only as much as its traceability. A plain attendance list says someone was in the room, not that adequate competence for their role was built. The record holds up through a few clear fields and a checkable work result.

Measure delivered
   |
Record fields:   who (role)   ->  what content
   |               when         ->  which system/context
   |               result       ->  artifact or check
   |
Adequacy justified  ->  audit-ready internal evidence

The lever is the last field. An attendance confirmation proves presence, an artifact proves application. Pointing the record at a concrete result, such as a built template, a solved exercise on a real case, or a short passed check, makes visible that competence was not only taught but applied. That is what separates a solid record from a mere formality.

A worked mini-example

An illustrative case in a 40-person department using AI tools daily. Target: all 40 people have a role-appropriate measure, documented with role, content, date and result. Survey after the first pass: 32 people with a complete entry including an artifact, 6 with attendance only and no result, 2 open. Coverage: 32 of 40 are solidly documented, that is 80 percent. The 6 attendances without a result and the 2 open cases are the concrete gap to close next. These numbers are a planning model, not legal advice. The point is measurability: once the record has clear fields per person, a coverage rate becomes visible, and the coverage rate turns into a concrete to-do list instead of a vague sense that training did happen.

Use cases by function

Function Task on the record Typical result
HR and L&D create and maintain the record training file with roles and results
Data protection and compliance assess adequacy documented justification of the measure
IT and information security provide the system link mapping of measure to systems
Department heads report roles and use complete audience
Executive management own the record approval and overview
Internal audit sample-check the record findings on gaps

Industries for which the record matters most

Demand rises with regulation and audit intensity. In finance and insurance and in health and pharma, training records are routine anyway, and the AI literacy record fits into existing documentation duties. In public administration, the clean, traceable record counts. In corporates with internal audit, the record is sample-checked, which forces completeness. In the mid-market, it is often about setting up a simple, maintainable trail at all, rather than being able to prove nothing in the end. The common denominator is that at some point someone can ask what was done, and the answer must be findable.

Term What it attests Relation to the literacy record
AI literacy record that an adequate measure took place, with a result the internal, audit-ready trail
Attendance confirmation that someone was present a building block, not a full record
Certificate that a provider's exam was passed can support evidence, but is not required
Reference / testimonial an overall assessment of a person different purpose, no measure trail
Skills matrix the team's current abilities a snapshot that complements the record

The most common mistake is equating an attendance confirmation or a provider certificate with proof of adequacy. Both can be part of it but do not replace your own complete documentation.

When a record fits, and when it degrades into a formality

It fits when a real measure took place and the company wants to record it traceably, with role, content, date and result. It degrades into a formality when only attendance is ticked off, without competence being applied or checked. A record without a proven result creates a false sense of security. So it pays to include a small, checkable artifact in the record from the start, rather than collecting signatures.

AI literacy record and the EU AI Act

Since 2 February 2025, Article 4 of the AI Act requires providers and deployers to ensure a sufficient level of AI literacy among staff. The rule prescribes no certificate but requires that the adequacy of measures is traceable if questioned. This is exactly where the literacy record fits. It can document implementation and serve as internal evidence. It does not guarantee automatic compliance and does not replace case-specific legal review. Whether the overall programme is adequate is assessed by the company itself.

Next step

Two ways, depending on where you are.

  • Book directly: Book a discovery call. 30 minutes, we look at your roles and measures and set the fields for a record that holds up.
  • Read along first: Enter your email and get the literacy record template with role, content, date and result field. No spam, unsubscribe anytime.

Build directive (Lovable): two side-by-side CTA cards (stacked on mobile). Card 1 = primary "Book a discovery call" button to https://cal.com/jamboula/ai-hackathon. Card 2 = email capture (<input type="email">, GDPR consent checkbox, double opt-in, submit to the lead list, inline success/error). Buttons carry a Phosphor icon (phosphoricons.com: CalendarCheck, EnvelopeSimple), hover/focus states via Motion (motion.dev, transform/opacity only), respect prefers-reduced-motion. This block also appears once higher up after the short definition.

FAQ

Is an AI literacy record legally required? A specific record in a fixed form is not prescribed. Article 4 requires adequate AI literacy measures and their traceability. Good internal documentation is the practical way to establish that traceability, but the form is left to the company.

Is an attendance confirmation enough as a record? It is a building block but rarely sufficient. An attendance confirmation proves presence, not applied competence. A solid record additionally points to role, content and a checkable result, such as an artifact or a short check.

What belongs in the record at minimum? Per person: role, date, content of the measure, link to systems and context of use, and a result field. The role link makes adequacy justifiable, the result field makes the record hold up.

How long should we keep the record? That depends on internal rules and the relevant legal situation. As orientation, keep records findable for as long as the measure might need to be justified. The specific retention period is settled by your own compliance.

Is this legal advice? No. Regulatory questions require review of the specific facts and current law by qualified counsel.

EU AI Act Article 4 · AI literacy · AI training · Practical transfer · In-house AI training · Copilot training

Sources and technical context

Discovery Call

Termin direkt buchen

30 Minuten, in denen wir Ausgangslage, Teamgröße und passende Challenges klären. Danach weißt du, welches Format zu euch passt.

cal.com · Corporathon Discovery Call

Der Kalender wird von cal.com geladen und setzt dafür eigene Cookies. Wir laden ihn erst, wenn du zustimmst oder ihn hier direkt anforderst.

Alternativ direkt auf cal.com buchen