The one AI Act deadline everyone hoped would slip — didn’t. It’s 2 August.

For most of this year, a lot of European businesses were quietly betting the AI Act’s literacy duty would get pushed back. The Digital Omnibus was supposed to be the reprieve.

It wasn’t. From 2 August 2026, national market-surveillance authorities gain the power to supervise Article 4 — the duty to ensure your people have "a sufficient level of AI literacy." The obligation itself has applied since 2 February 2025. What arrives on 2 August is the enforcement.

What Article 4 actually says

It’s short, and deliberately not prescriptive. Providers and deployers of AI systems must "take measures to ensure, to their best extent, a sufficient level of AI literacy" among the staff who use those systems — proportionate to the organisation’s size and the context of use.

Two things are worth reading carefully:

• It’s an obligation of effort, not a certificate. Nobody hands you a pass/fail exam. But "we did nothing" is not a defensible position for an organisation that puts AI tools in front of its people.
• It applies to deployers, not just AI vendors. If you’ve rolled out Copilot, ChatGPT Enterprise, or any AI feature inside the tools you already own — that’s you.

What the Digital Omnibus did and didn’t do

The proposal circulating in Brussels would soften *how* the duty is framed — shifting some weight toward the Commission and member states to "encourage" literacy rather than placing it squarely on deployers.

But two facts matter more than the framing debate:

1. It does not move the 2 August enforcement date.
2. It is a proposal, not law. Planning your year around a softening that may or may not pass — and won’t arrive before the deadline anyway — isn’t a plan. It’s a hope.

This isn’t a compliance project. It’s an adoption problem wearing a deadline.

Here’s the part most vendors won’t tell you, because they’d rather sell you a compliance module.

If your people genuinely know how to use the AI you’ve already paid for — what it’s good at, where it’s wrong, what not to put into it, how to check its output — you have already met the spirit of Article 4. The literacy foundation is a by-product of doing adoption properly, not a separate box to tick.

And the inverse is the uncomfortable one: if you can’t show your staff are AI-literate, it’s usually because they aren’t *using* the AI — which means you’re also carrying the waste. The same gap shows up on two different reports: the auditor’s and the CFO’s.

What "sufficient AI literacy" looks like in practice

Not a lunch-and-learn nobody remembers. Something you could actually show:

• Role-based — finance, ops and leadership don’t need the same things.
• Documented — who was trained, on what, when. An auditable record, not a vibe.
• Practical — grounded in the actual tools your people touch, the actual data they must never paste, the actual judgement calls the AI can’t make for them.

That’s the same thing that makes AI pay back. Trained, confident people use the tool; untrained people quietly do the work by hand and the licence sits idle.

Where we stop

We build that literacy foundation — the role-based enablement and the record that it happened. We do not give you legal or compliance advice on the AI Act; that’s a job for a qualified partner, and we’ll happily point you to one.

The deadline is the reason to act now. The reason it’s *worth* acting is that the same work fixes the adoption gap you’re already paying for.

You don’t need a compliance project. You need your people actually able to use the AI you already bought — and a record that they can.

The honest first step is small: find out where you actually stand.