If you run a team in Dubai, Abu Dhabi, Riyadh, or Jeddah, you’ve probably noticed AI literacy shifting from a conference topic to something with a policy document attached. It’s worth understanding what’s actually been decided, because the second-order effects land on private employers even though the mandates mostly don’t.
What has actually happened
Saudi Arabia designated 2026 its Year of Artificial Intelligence, and the substance behind the branding is unusual in its scale:
- SDAIA’s SAMAI initiative has trained over one million Saudi citizens in AI skills — around nine percent of the working-age population. Very few countries anywhere have moved a population-level number like that.
- SAMAI 2 extends the programme across the public sector, launched in partnership with eleven government ministries.
- A mandatory National Cross-Disciplinary Curriculum for Data and AI now runs across all undergraduate disciplines at Saudi universities — not just computer science. Law graduates, business graduates, medical graduates.
- Around 11,000 AI specialists have been trained through dedicated tracks.
The UAE has built AI literacy into national strategy rather than treating it as a training programme:
- The National AI Strategy 2031 targets a substantial rise in AI’s contribution to national GDP — from roughly 9% today toward a far higher share by 2031, with hundreds of billions of dirhams in projected value.
- The UAE AI Charter places literacy alongside governance and innovation as a pillar rather than an afterthought.
- AI for All, developed with Google, is delivering AI fundamentals training throughout 2026.
- AI is now a mandatory school subject, with curriculum rolled out from the 2025–26 academic year across all public-school stages, covering concepts, applications, ethics, and responsible use.
- One Million Prompters targets practical, applied skills at population scale.
Read those two lists together and the pattern is clear. This isn’t a pilot or a pledge. Both countries have decided AI capability is national infrastructure and are funding it accordingly.
What it does and doesn’t require of you
Be precise here, because a lot of vendor marketing in the region is currently blurring this line.
These mandates don’t apply to private employers. The compulsion is in education and the public sector — school curricula, university requirements, government ministries. If you run a private company in the UAE or Saudi Arabia, nobody is going to audit your staff’s AI training.
What changes anyway is the baseline. Three effects, in rough order of how soon you’ll feel them:
1. New hires arrive different. A Saudi graduate entering the workforce now has completed a mandatory data and AI curriculum regardless of what they studied. Emirati students have had AI as a school subject. Within a few intake cycles, “familiar with AI tools” stops being a differentiator on a CV and becomes an assumption — while your existing staff, who went through university before any of this, quietly become the ones who need support.
2. Government-linked work starts asking. Organisations bidding for public-sector work in either country increasingly encounter capability questions. Aligning your training with national priorities is also, in plain commercial terms, a good relationship to have.
3. Your competitors get free training. SAMAI and AI for All are national programmes reaching enormous numbers of people at no cost to employers. If a competitor’s staff are using them and yours aren’t, that gap costs you nothing to close and something real to ignore.
The trap: mistaking awareness for capability
Here’s where organisations in the region are most likely to waste money over the next eighteen months.
National programmes are very good at the awareness layer — what AI is, what a prompt is, what the ethical considerations are, how not to embarrass yourself or your employer. That layer is being delivered at population scale, competently, and largely for free.
So if your corporate training plan is also the awareness layer, you are paying for something your government is already giving your staff. Worse, you’ll conclude that training didn’t produce results — because awareness doesn’t produce results. It produces people who know the vocabulary.
The gap that remains, and that nobody is filling at scale, is operational capability: can this person hand a genuine multi-step piece of work to an agent, scope it safely, and verify what comes back well enough to put their name on it?
That’s a different skill. It’s closer to learning to delegate to a new colleague than to learning a software feature. It’s the difference between someone who can ask an AI to draft an email and someone who can hand it the folder of regional exports and get back a reconciled report they’d defend in a meeting. What that actually looks like.
Why Arabic matters more than the brochures admit
Both national programmes are delivered with Arabic as a first-class language, and that sets an expectation your corporate training will be measured against.
A great deal of English-language AI training assumes an English-language workplace. In Gulf offices, that assumption breaks constantly and in a specific way: the source material is Arabic — supplier emails, contracts, government correspondence, customer messages — while the deliverable is often English. Or the reverse. Or both in the same document.
Training that only ever demonstrates English-to-English work leaves people to discover the mixed-language reality on their own, which is where most of the real difficulty lives. If you’re evaluating a programme, that’s the question to ask: does it teach the work as it’s actually done here, in both languages, or is it a translated version of a Western curriculum? More on how Claude handles Arabic in practice.
What to actually do about it
1. Stop paying for the awareness layer. Point staff at the free national programmes for fundamentals — SAMAI, AI for All, One Million Prompters. They’re good, they’re funded, and they’re aimed at exactly this. Spend your budget on the layer above.
2. Find where the real work is stuck. Not “who wants AI training” — that gets you the enthusiasts. Ask who spends the most hours moving information between files, chasing documents, and rebuilding the same report every month. In most organisations that’s operations, finance, HR, and analysts, not the technology team. How to pick that cohort.
3. Teach delegation and verification, not prompting. The durable skills are scoping a task well, knowing what an agent should and shouldn’t be trusted with, and checking output before it leaves the building. Prompt phrasing is the part that changes every six months; judgement is the part that doesn’t.
4. Make capability provable. This is where the policy environment genuinely helps you. In a region where governments are counting trained citizens and publishing the numbers, “we did some AI training” is a weak claim, and named people, named tasks, and before-and-after numbers are a strong one. It survives the person who championed it leaving, it holds up in a tender document, and it distinguishes people who completed something from people who attended something.
5. Do it in both languages, deliberately. Not as a translation exercise at the end — as the way the material is built. If your Arabic-speaking staff are consuming a worse version of the training than your English-speaking staff, you’ll see it in adoption, and you’ll misdiagnose it as reluctance.
The bigger picture
There’s a version of the next few years where AI capability in the Gulf follows the same path as English proficiency did — starting as a differentiator, becoming an expectation, and ending as something nobody remarks on except when it’s absent.
The national programmes are what make that trajectory likely rather than speculative. When a country trains nine percent of its working-age population in a single initiative, it isn’t running a pilot; it’s resetting the floor.
For an employer, the useful conclusion isn’t urgency for its own sake. It’s that the awareness problem is being solved for you, at scale, by someone else — and that the thing left for you to solve is narrower, harder, and considerably more valuable: turning people who know what AI is into people who can hand it real work and stand behind the result.
If you’re ready to start, the Gulf-specific adoption guide covers what this looks like day to day for a Dubai or Riyadh team.