Fifty days. That is the gap between PwC’s August warning and Visa’s October protocol launch. Three days later, McKinsey published the playbook. In that compressed window, the entire payments and commerce infrastructure admitted it was playing catch-up to a market already in motion.
If you are a retail business leader in the Middle East reading this in December 2025, you are reading documents about a transition that has already finished elsewhere. The question is not whether agentic commerce will happen. It is whether you will be a participant or a spectator in your own business.
AI agents are not a future shopping tool. They are already routing purchasing decisions away from your website, your customer relationships and your control. In August, PwC documented a 1,200% surge in AI-originated retail traffic. Visa then launched a protocol to maintain visibility over transactions it could no longer see. McKinsey followed with the playbook.
This is not innovation. This is institutional catch-up.
For GCC retail, the timeline is different but the direction is identical. You have perhaps eighteen months before agents become the primary interface between consumers and your inventory. Not the primary discovery channel. The primary interface. Your customer will tell their agent what they need. The agent will decide where to buy it. You will fulfil the order, if you are lucky.
An AI agent shopping for athletic wear will compare your offering against two hundred competitors in three seconds. It will weight price, delivery speed and reviews using parameters you did not choose. Your brand premium, your loyalty programme, your customer service reputation. All secondary to the algorithm’s optimisation function.
You cannot outbrand an agent. You cannot emotionally persuade a machine.
Become indispensable at the infrastructure level. Structure your product data so machines can read it. Make it available to systems you do not control. If an agent cannot access your inventory in real time, you become invisible.
Build or partner for API access to agent platforms. The retailers winning are not the ones building their own agents. They are the ones woven into the dominant platforms. Abandon the fantasy that your website traffic will sustain your business.
Your margins will compress. Price-hunting agents will turn your fast-moving products into commodities. You will compete on cost and speed. This is not temporary. This is permanent. Your unit economics need to shift now.
Consumers in Germany and Japan prefer to control their own money. They distrust algorithms. In the Middle East, digital adoption varies. Some cultures value autonomy more than efficiency. An agent spending your money is not intuitive everywhere. This buys you time. You can shape how agents behave in your region before they are set in stone.
Partner with agent platforms on terms favourable to regional merchants. Right now, these platforms need merchant data and fulfilment partners. They need you more than you need them. Only until they do not. Move fast.
Stop defending the old model. Start building the new one. Not in two years. Now. Move your best people into agent integration. Accept that you no longer own the customer relationship. What keeps you alive is a fast, reliable operation and clean data. Not brand strength. Not loyalty programmes.
For twenty years, you have invested in direct customer relationships. You have built websites, trained teams, created loyalty programmes. Much of this dies in an agent-driven market. What you built only matters if agents trust you to deliver. To fill orders fast. To send accurate data. You are becoming a back-office operation.
Infrastructure providers in previous technology transitions often became more profitable than the visible layer above them. You will make that transition before your competitors do, or you will not.
Agent platforms are American or Chinese. European players are building sovereignty frameworks. The Middle East has not yet chosen its position. Shape how agents operate in your markets. Only if you move fast enough to have that conversation before the platforms set their default behaviours.
For GCC retail business leaders, the work is concrete. In the next sixty days, know your data architecture. Know which API platforms exist. Know which agent platforms your customers will use first. Negotiate before you become optional.
The executives who will lead retail in five years will not be those with the best loyalty programmes. They will be those who understood, in 2025, that their customer is no longer a person but an algorithm.
The transition is here. Lead it or follow it.
Those who do neither will disappear.