Agentic commerce could move USD 3-5 tn globally by 2030 — Egypt’s merchants are interested but largely unfamiliar with what it means

🤖🛒 “An identity and a wallet.” That was Cloudflare’s announcement in August, when the cloud technology platform launched two new services designed to give AI agents on its platform a stable identity and the ability to make purchases online securely — within limits set by the humans who created them. The implication? Merchants and buyers could, for the first time, rely on digital commerce agents with confidence, knowing who they’re dealing with and that every transaction is secure at every stage. Right?

In practice, the announcement generated more skepticism than celebration — as tends to happen with AI. The development added fresh fuel to a debate already running through the payments sector: are we seeing the foundations of trust being built at the same pace as the infrastructure itself?

How it works: Agentic commerce runs on AI-powered systems that execute buying and selling autonomously, designing workflows using available tools. Where traditional e-commerce requires the buyer to manually search for products, compare options, read reviews, and complete payment step by step, agentic commerce shifts most of that work to agents who shop on the user’s behalf.

AI in stages: At one point or another, we’ve all searched for a product and watched Google’s AI summaries immediately surface recommendations. AI’s role in search and shopping is now foundational, and it's developing across three stages:

• Stage one — discovery assistant: Many consumers already use AI to find products and surface recommendations. Around 50% of consumers rely on it when searching for products, according to a recent McKinsey report;
• Stage two — agentic interface: Here, AI agents suggest options and facilitate transactions through conversational interfaces, while the consumer retains oversight and final approval — which is the core of the agentic commerce concept. This is where protocols like OpenAI’s Agentic Commerce Protocol and Google’s Universal Commerce Protocol have emerged, with the tech giant partnering with commerce platforms like Shopify to enable AI agents to handle everything from product discovery through to payment and beyond;
• Stage three — autonomous purchasing: Agents gain significant independence. The vision here involves completing purchases automatically on the consumer’s behalf — if your fridge runs low on dairy, the smart app orders it. This scenario focuses on routine and essential goods, giving consumers a sense of ease and control over repetitive daily tasks rather than having to handle them manually.

Cloudflare isn’t alone in this space. Visa has developed its Trusted Agent Protocol, and Mastercard has launched Agent Pay — services that give these systems a verifiable identity, a defined spending limit, and an auditable transaction record that both merchants and consumers can rely on.

The openings

Agentic commerce could handle transactions worth USD 3-5 tn globally by 2030, according to McKinsey estimates, with the US consumer retail sector alone projected to account for around USD 1 tn — driven by growing reliance on AI agents for product search, comparison, and purchase decisions on behalf of consumers.

Inventory management and demand forecasting: The global retail sector loses around USD 1.73 tn annually due to inventory errors — whether stockouts or overstocking — according to IHL research. AI agents can be deployed to analyze current sales data, shopper movement patterns, and seasonal trends to automatically reorder stock or reprice slow-moving items, keeping supply chains efficient and preventing losses from poor planning.

Dynamic pricing and automated negotiation: These systems can monitor competitor pricing around the clock and adjust product prices in real time to maintain competitiveness and maximize margins. Their capabilities extend into B2B, through automated negotiation with suppliers based on supply and demand data.

Hyper-personalization: The system tracks customer behavior across a digital storefront and proactively surfaces offers tailored to individual preferences while also managing returns and handling complex inquiries.

Fraud reduction: Some argue that AI agents represent an advanced layer of protection capable of analyzing transactions in real time to flag suspicious behavior — whether in e-commerce fraud attempts or unusual patterns at self-checkout terminals in physical stores.

Where does Egypt (and the region) stand?

Does the Egyptian merchant care about AI agents? AI technologies such as personalization and autonomous transactions have attracted interest among local retail and commerce players at a rate of 55.7%, according to a Fast Company survey conducted in partnership with Visa, which covered 300 participants from the UAE and Saudi Arabia and 200 from Egypt.

However, while the technology is advancing rapidly, the sector still needs to fully grasp what it means to rely on autonomous agents. Nearly three-quarters of Egyptian survey respondents remain unfamiliar with the concept of agentic commerce.

Saudi Arabia: In the Saudi market, major e-commerce platforms like Salla have moved quickly to integrate AI tools, enabling merchants to automate key operations — including customer communication, order tracking, and abandoned cart reminders — reducing the need for direct human intervention and improving operational efficiency.

Looking ahead: Despite relatively low awareness of what the technology actually means, more than 83% of Egyptian companies plan to invest in agentic commerce within the next two years. More than 70% expect agent-based technologies to have a noticeable impact — even if modest — on their sectors over the same period. Around six in 10 companies in the UAE and Saudi Arabia expressed similar interest, according to the survey.

The potential roadblocks

Security, governance, and more: Among the most pressing obstacles, data privacy and security top the list for Visa survey participants. AI agents process vast amounts of sensitive data — like payment details and personal information — which demands investment in robust encryption and compliance systems. Merchants also worry about unclear ROI alongside governance and trust challenges that go beyond technological limitations and skills gaps. This is understandable: at a time when Egyptian users are already contending with various forms of digital fraud, trusting an AI agent to shop and pay on their behalf is a significant ask.

The tech world has also seen a number of recent incidents where AI models have gone off-script, breaching tightly controlled systems — pushing companies like OpenAI to suspend training on some of their models. That same behavior could surface in an AI agent making a purchase: what ensures that the data it holds and uses to make buying decisions is secure?

What comes next: Retail companies using AI and machine learning have achieved 2.3x sales growth and 2.5x earnings growth compared to their peers, according to the IHL report. However, the gap between deploying AI tools in the background and handing effective control to an agent making semi-autonomous decisions remains significant — particularly given the shift toward slowing the rapid development of AI tools out of concern for consequences that could prove difficult to manage. That may mean a meaningful slowdown in the technology’s path toward full autonomy in the near term.