Als nächstesNächster Leitfaden
Data Governance Requirements in EU AI Act Article 10
Gesellschaft
Gesellschaftsführer
Article 4 of the EU AI Act requires providers and deployers to take measures supporting AI literacy among staff and other people operating or using AI on their behalf.
The 2026 amendment tailors the obligation to people and context and expressly says it does not require guaranteeing a specific literacy level for every individual.
Article 4 is a binding obligation in Regulation (EU) 2024/1689, the EU AI Act. In its original form, it required providers and deployers to take measures to ensure a sufficient level of AI literacy. Regulation (EU) 2026/1744, in force from 27 July 2026, amended the provision. The current wording asks providers and deployers to take measures to support the development of AI literacy among their staff and other people who deal with operation and use on their behalf. Measures must take account of people’s technical knowledge, experience, education and training, the system’s use context, and the people or groups on whom it will be used. The amendment explicitly states that the obligation does not require a provider or deployer to guarantee any specific literacy level for any individual. This is not an exemption from taking measures: organisations should still identify who needs what understanding to use, oversee or support a system responsibly. Article 4 applies to providers and deployers, not only employers or high-risk systems. A provider develops or has a system developed and places it on the market or puts it into service under its name; a deployer uses a system under its authority, subject to the Act’s definition and exclusions. Article 4 applied from 2 February 2025 under the Act’s phased timetable. After the 2026 amendment, the Commission and Member States must support and facilitate compliance efforts, especially for SMEs, including by publishing practical examples on the single information platform. The law does not prescribe one standard course, exam, number of hours or certificate. A sensible programme is role-based and documented, but examples and recommendations should not be confused with a binding universal curriculum.
Sowohl katastrophale als auch alltägliche Schäden durch KI hängen davon ab, wer die Risiken versteht und wer handeln kann.
Die öffentliche und berufliche Bildung bestimmt, ob eine starke Sicherheitspolitik politisch möglich ist.
Klare Erklärungen reduzieren die Vereinnahmung durch Hype, Labor-PR und vages Ethik-Theater.
The 2026 amendment shifts emphasis from guaranteeing an individual threshold to taking measures that support development. The Commission and Member States are tasked with support, including practical examples, while providers and deployers retain the measure-taking duty. Teams should check the live EU AI Act platform for official examples and distinguish them from nonbinding recommendations or commercial training products. Teams should track official examples as they appear, then review whether measures still fit new roles, systems and affected groups. Record the rationale for tailoring without turning guidance into a fixed syllabus.
A provider maps staff roles from model developers to support teams and offers role-appropriate guidance on system limits and escalation.
A deployer using an AI scheduling tool trains operators on the tool’s purpose, intended use and when a human must intervene.
A small business evaluates existing staff knowledge and chooses proportionate learning resources rather than promising every employee a fixed certificate.
A public service explains the system to frontline staff in light of affected groups and the context in which decisions are made.
Das existentielle Risiko wird als Science-Fiction behandelt, während sich die Fähigkeiten verstärken.
Verwechslung von Oberflächenproduktsicherheit mit Ausrichtung unter hoher Autonomie.
Nicht-englischsprachigen und nicht fachkundigen Zielgruppen stehen nur Quellen von geringer Qualität zur Verfügung.
Separate Risiken für Produktschäden, Missbrauch und Kontrollverlust/Fehlausrichtung.
Fragen Sie, welche Beweise Ihre Sicht auf Zeitpläne und Schweregrad ändern würden.
Bevorzugen Sie Primärquellen und konkrete Bewertungen gegenüber Marketingaussagen.
Identifizieren Sie einen Aktionspfad: Karriere, Politik, Finanzierung oder Fähigkeiten – nicht nur Bewusstsein.
Free newsletter
Three verified AI stories every weekday morning, written in plain English. Free forever, no ads.
One email each weekday. Unsubscribe in one click. We never sell or share your address.
Test yourself
Instant feedback on every answer, and a shareable certificate with a verifiable ID once you pass a course.
Support free AI education. AI Understanding is a 501(c)(3) nonprofit — no ads, no paywall, ever. Make a donation
Article 4 of the EU AI Act requires providers and deployers to take measures supporting AI literacy among staff and other people operating or using AI on their behalf. The 2026 amendment tailors the obligation to people and context and expressly says it does not require guaranteeing a specific literacy level for every individual.
After amendment, Article 4 requires measures supporting development of literacy among staff and other people handling operation or use on the organisation’s behalf.
The text covers staff and other persons who deal with operation and use on behalf of providers or deployers.
The amended article expressly disclaims a guarantee of a specific literacy level for any individual.
The Act’s phased application dates put Articles 4 and 5 in application from 2 February 2025.
Article 4(2) requires public support and practical examples, while providers and deployers retain their duty.
Lerne weiter
Weitere Leitfäden zu diesem Thema ausgewählt
Als nächstesNächster Leitfaden
Data Governance Requirements in EU AI Act Article 10
Gesellschaft