EU AI Act

Article 4: AI Literacy

What it means for your organization, your team, and your career. And how the Protocol helps you comply.

Enforcement Timeline

2 February 2025 — Article 4 applies

The obligation to ensure AI literacy of staff already applies. Organizations must begin taking measures. EU AI Act (Regulation 2024/1689) ↗

2 August 2025 — National enforcement regimes due

Member States must designate supervising authorities and lay down their penalty rules; the general-purpose AI model obligations also start applying.

2 August 2026 — Enforcement begins

The Act becomes generally applicable: the Article 50 transparency obligations (AI disclosure, deepfake labeling) take effect, market surveillance starts, and national authorities can enforce the Article 4 duty that has applied since 2025. This date was not deferred by the Digital Omnibus. European Commission — AI Act overview ↗

2 December 2027 — High-risk obligations (deferred)

What the EU Digital Omnibus (agreed May 2026) actually deferred: the obligations for standalone high-risk systems under Annex III move from August 2026 to December 2027, and Annex I embedded systems to August 2028. The deferral does not touch Article 4 or the transparency rules. Full timeline with sources: EU AI Act Article 4 — what organizations need to know.

What Article 4 Requires

Who is covered?

All providers and deployers of AI systems. This includes any company that develops, distributes, or uses AI systems — which in 2026 means practically every organization.

What must you do?

Take measures to support the development of AI literacy among all staff and external service providers involved in planning, implementing, or using AI systems. Training must account for technical knowledge, experience, and context of use.

Updated 3 August 2026. Article 4 originally required you toensure a “sufficient level of AI literacy”. Regulation (EU) 2026/1744 (in force 27 July 2026) changed this totaking measures to support the development of it — an effort obligation rather than an outcome one. The duty and its dates did not move; the standard did. You are expected to provide and evidence the opportunity, not to guarantee every person reached a threshold.

How is compliance measured?

There is no one-size-fits-all standard. The AI Office does not impose mandatory specific training programs. But you need to demonstrate that training was provided, personalized to roles, and documented.

What are the penalties?

Article 4 has no dedicated fine tier in the Act itself — breaches are sanctioned through the national penalty regimes Member States adopted, which authorities can apply from 2 August 2026. The neighbouring transparency obligations (Article 50, also applicable from August 2026) carry EU-level fines up to EUR 15 million or 3% of global annual turnover. In practice the earlier cost is indirect: AI literacy evidence is requested in audits, procurement, and incident investigations.

How the Protocol Supports Article 4 Readiness

• Personalized to role and context: The Protocol generates training based on each individual's specific role, industry, tasks, and current AI readiness level.
• Structured and documented: Every Protocol is a 30 or 60-day structured plan with daily tasks, milestones, and measurable outcomes. Fully auditable.
• Evidence-based methodology: Built on a curated evidence base of real AI deployments across 15 sectors. Not theoretical — based on actual workforce impact data.
• Progressive skill levels: Four maturity levels (AI-Curious to AI-Native) ensure training matches each person's starting point and grows with them.
• PDF deliverable for compliance records: Every Protocol is downloadable as a PDF — ready for compliance documentation, HR records, and audit trails.

Rolling this out for a team?

The Protocol is sold per person — each individual gets a plan personalized to their own role, tasks, and AI readiness, and each finished Protocol is a documented, downloadable record of structured training. We don't currently sell company audit packs or volume licensing. If your organization has specific needs, write to us.