What is the EU AI Act Article 4?

Article 4 requires providers and deployers of AI systems to ensure their staff have a sufficient level of AI literacy. It has applied since 2 February 2025. Under the May 2026 Digital Omnibus provisional agreement it is being reframed as a duty to support the development of AI literacy, enforceable by national authorities from August 2026.

Did the August 2026 high-risk deadline change?

Yes. Under the Digital Omnibus (provisional agreement, 7 May 2026), high-risk obligations for Annex III systems were deferred from 2 August 2026 to 2 December 2027, and Annex I from 2 August 2027 to 2 August 2028. The Article 4 AI-literacy duty, in force since February 2025, was not delayed.

Am I a provider or a deployer under the Act?

Most organisations are deployers: you use an AI system in your operations. The moment staff use tools like ChatGPT, Copilot or Claude for work, Article 4's literacy duty applies to you.

The EU AI Act & Article 4, in Plain English: a Guide for HR Leaders

Q: Does the EU AI Act apply to UK companies?

It can. The Act has extraterritorial reach: it applies to organisations outside the EU, including the UK, if they provide AI systems into the EU market or if the output of their AI is used in the EU.

There's been a lot of noise, and several shifting deadlines, around the EU AI Act. Here's the honest version: one big deadline just moved, but the obligation most likely to affect your people is already in force.

The headline, up front

The much-quoted 2 August 2026 deadline for high-risk AI systems has been deferred to December 2027 under the Digital Omnibus. But the Article 4 AI-literacy obligation has applied since 2 February 2025 and becomes enforceable from August 2026. Literacy is the live, near-term duty, not the thing that got delayed.

What is the EU AI Act?

The EU AI Act is the world's first comprehensive law regulating artificial intelligence. It entered into force on 1 August 2024 and applies in stages, using a risk-based approach: the more potential for harm, the heavier the obligations.

Unacceptable risk

Banned: e.g. social scoring, manipulative systems. Prohibited since Feb 2025.

High risk

Allowed but heavily regulated: e.g. AI in hiring, credit, education. The bulk of the obligations.

Limited risk

Transparency duties: e.g. telling people they're talking to a bot, labelling AI content.

Minimal risk

Most everyday AI (spam filters, etc.): no specific obligations.

Does it apply to UK companies?

Potentially, yes. The Act has extraterritorial reach: it can apply to organisations outside the EU (including the UK) if they provide AI systems into the EU market, or if the output of their AI is used in the EU. Many UK firms with EU customers, staff or operations are in scope.

Article 4: the AI-literacy obligation

Article 4 is the part most relevant to HR, because it's about your workforce, not your software.

Article 4, in essence

Providers and deployers of AI systems must ensure a sufficient level of AI literacy among their staff and others operating AI on their behalf, taking into account those people's knowledge and training, the context of use, and the people it's used on.

“Provider” vs “deployer”: which are you?

Provider: you develop an AI system (or have one built) and put it on the market under your name.
Deployer: you use an AI system in your operations. Almost every organisation is a deployer the moment staff use ChatGPT, Copilot or Claude for work.

What changed in 2026

Under the Digital Omnibus (provisional political agreement, 7 May 2026, pending formal adoption), Article 4 is being softened: from a duty to “ensure a sufficient level” of literacy to a duty to “take measures to support the development of” literacy.

Don't mistake softer for gone

The obligation still exists, still applies to deployers, and becomes enforceable by national authorities from August 2026. “Support the development of literacy” still means delivering real, role-appropriate training, and being able to evidence it.

The timeline that matters

1 Aug 2024Done

EU AI Act enters into force.

2 Feb 2025In force

Prohibited practices banned. Article 4 AI-literacy obligation begins to apply.

2 Aug 2025In force

Rules for general-purpose AI (GPAI) models and the governance/penalties framework take effect.

Aug 2026Next up

Enforcement of Article 4 begins, via national authorities. The original high-risk deadline fell here, but has moved (below).

2 Dec 2026Next up

Tightened deadline for transparency / labelling of AI-generated content.

2 Dec 2027Deferred

High-risk obligations (Annex III, use-based): moved here from Aug 2026.

2 Aug 2028Deferred

High-risk obligations for regulated products (Annex I): moved here from Aug 2027.

What companies need to do now

Regardless of the high-risk delay, these are live today because Article 4 is already in force:

Action	Why it matters now
Inventory AI in use	Including shadow AI. You can't govern or train for tools you can't see.
Baseline literacy training	Role-appropriate, hands-on. The core of the Article 4 duty.
A clear safe-use policy	What data can and can't go into AI tools: your first line against leaks.
Name an owner	One accountable person for AI literacy and governance.
Keep evidence	Dates, attendance, policy sign-off; enforcement lands August 2026.

What to do before the key dates

Before August 2026 (literacy enforcement): role-appropriate training delivered and documented, a safe-use policy live, and an accountable owner in place. Keep the evidence in one place: here's a free fill-in Article 4 record template.
Before December 2026 (transparency): if you generate AI content at scale, make sure it's labelled as required.
Through 2027 (high-risk): if you deploy AI in hiring, credit or education, use the extra time to map those systems and prepare; the delay is breathing room, not a reprieve.

The Zero-Hype take

The deadline that got the headlines moved. The obligation that touches your workforce did not. Treat AI literacy as box-ticking panic and you'll waste money; use it to build a genuinely capable, confident workforce and a compliance duty becomes a productivity advantage.

See where you stand in 2 minutes

Take the free AI-Fluent Workforce scorecard, or book a no-pitch call.

Take the AI fluency check Book a call

Sources & further reading

European Commission: Regulatory framework on AI & AI Literacy Q&A
Article 4: AI literacy & Implementation timeline
Council of the EU: Agreement to simplify the rules (7 May 2026)

This guide is for general education and does not constitute legal advice. Verify obligations specific to your organisation with qualified counsel.