- The UAE has set a target for 50% of government services to run on Agentic AI within roughly two years, and Dubai has extended the expectation across the private sector.
- Agentic AI means autonomous systems that decide and act across multi-step workflows — not chatbots.
- Most organisations lack a baseline, a prioritised roadmap, and board-ready reporting against the deadline.
- The winning move is to baseline honestly, sequence for commercial value, and report progress monthly.
In 2026 the United Arab Emirates issued the most ambitious artificial-intelligence directive any government has put forward: a target for 50% of federal government services to move to Agentic AI within roughly two years. Dubai went further, signalling through the Dubai Chamber of Commerce that the private sector is expected to follow. For any business operating in the Emirates, the question is no longer whether to adopt AI, but how quickly you can show real, defensible progress.
This article explains what the mandate actually requires, why most companies are not ready, and the practical sequence that turns a daunting deadline into a manageable programme — the same sequence I use with clients across telecom, financial services and healthcare in Dubai and the wider GCC.
What "Agentic AI" actually means
The headline figure — 50% — matters less than the word in front of it. Agentic AI is not a chatbot bolted onto a website, and it is not a single model answering a single question. It is software that operates autonomously: it analyses a situation, decides what to do, and executes multi-step work, calling on tools and data along the way, with human oversight at the points that carry risk.
In practice that means moving real operational workload — claims processing, customer resolution, document handling, compliance checks — onto systems that act, not just advise. That is a materially higher bar than the generative-AI pilots most organisations have run so far, and it is the bar the mandate is set against.
Why the mandate exists now
The UAE has spent a decade positioning itself as a global technology hub, from appointing the world's first Minister of State for Artificial Intelligence to its national AI strategy. The Agentic AI mandate is the logical next step: a forcing function that converts ambition into deployed systems on a fixed timeline. For the government it accelerates service delivery; for the private sector it sets a competitive baseline that laggards will struggle to meet after the fact.
The companies that start now will be in production before the deadline. The ones that wait will still be writing strategy decks while their competitors report progress.
Why most companies are not ready
The gap is rarely the technology — capable models and tooling are widely available. The gap is organisational. In nearly every business I assess, three things are missing:
- No baseline. Leadership cannot say what share of operations is "agentic" today, because no one has defined or measured it.
- No prioritised roadmap. Teams debate use cases in the abstract while the clock runs, with no agreed sequence tied to value and feasibility.
- Nothing to report. When the board or the Chamber asks for progress, there is no credible, repeatable way to show it.
The result is months of motion without momentum. Pilots get built to prove a vendor's point rather than to move a number the CFO cares about, and the mandate deadline gets closer with little to show.
Where to start: a three-move sequence
You do not need to solve the entire mandate at once. You need a defensible baseline, a sequenced roadmap, and a reporting rhythm. In practice that is three moves:
1. Baseline honestly
Establish where you stand today against the mandate's expectations — which processes already use AI, which are candidates, and what your data and governance can realistically support. An honest baseline is uncomfortable but invaluable: it ends the abstract debate and gives leadership a real starting line.
2. Sequence for value
Prioritise the moves that are both compliant and commercially worthwhile, so early wins fund and de-risk the harder ones. A well-sequenced roadmap front-loads provable results — a single agentic use case in production beats five stalled pilots every time.
3. Report monthly
Put a monthly reporting cadence in place that gives your board and the Dubai Chamber a clear, defensible view of progress, gaps and risk. Reporting is not bureaucracy here; it is the evidence that you are on track, and it keeps the programme honest month to month.
What happens if you wait
Late movers face a worse version of the same work under time pressure, plus a competitive disadvantage as early adopters compound their gains. There is a recovery path — a compressed discovery-to-deployment sprint can get a first agentic use case live in six to eight weeks — but it is more expensive and more stressful than starting deliberately now.
How I help
I work with leadership teams across the UAE and GCC to do exactly this: a readiness baseline, a sequenced roadmap to the two-year target, and monthly mandate-compliance reporting designed for the board and the Dubai Chamber — with an Accelerated Mandate Catch-Up option for organisations that are already behind. The goal is always the same: real systems in production, moving real numbers, with evidence you can stand behind.