Agentic Commerce: When AI Buys and Sells on Your Behalf

Imagine you need new running shoes. Instead of browsing dozens of stores, comparing prices, and reading reviews, you tell your AI agent: "Find me comfortable running shoes under $100." The agent searches every retailer, compares prices and reviews, waits for the best discount, and completes the purchase automatically.

This is not science fiction. This is agentic commerce, and it is happening right now.

At the National Retail Federation (NRF) conference in early 2026, a live audience poll revealed that 75% of attendees were either currently implementing or actively planning agentic commerce initiatives. Stripe launched its Agentic Commerce Suite, Google announced new open standards, and major brands like Etsy, Coach, and Kate Spade have already onboarded.

What Is Agentic Commerce?

Agentic commerce is a new business model where AI agents act as autonomous intermediaries between buyers and sellers. These agents go beyond recommendations — they search, compare, negotiate, and complete purchases independently within user-defined permissions.

Here is how it differs from current AI shopping tools:

How Agentic Commerce Works Under the Hood

Shared Payment Tokens (SPTs)

One of the most significant innovations is Stripe's Shared Payment Tokens. These tokens allow an AI agent to initiate payments using a buyer's saved payment method — without exposing actual payment details.

Each token is precisely scoped:

• Specific seller: The agent can only purchase from designated merchants
• Time limit: Token expires after a defined period
• Spending cap: Maximum spend that cannot be exceeded
• Full observability: Every transaction is trackable in real time

The Agentic Commerce Protocol

Stripe announced an open protocol endorsed by over 25 ecosystem partners including Salesforce, Squarespace, and PwC. This protocol standardizes how AI agents communicate with merchants, enabling seamless interoperability across different platforms.

Why MENA Is Primed for Agentic Commerce

The Middle East and North Africa region has ideal conditions for adopting this model:

1. Young, Digital-First Population

Over 60% of the population is under 35, with some of the highest smartphone and social media usage rates globally. This demographic is ready to delegate purchasing decisions to smart agents.

2. Accelerating Adoption Rates

In the UAE, 70% of consumers already use AI tools while shopping — a 44% increase from last year. And 80% of organizations plan to deploy AI agents by 2027.

3. Government-Led Innovation

From Saudi Vision 2030 to the UAE AI Strategy, governments across the region are providing supportive regulatory frameworks and advanced digital infrastructure.

How Major Brands Are Preparing

URBN's Incremental Approach

Rather than overhauling their entire product catalog at once, URBN (Urban Outfitters, Anthropologie, Free People) took a gradual approach. CIO Rob Frieman explains: "We focused on our most popular products and high-value use cases first" — starting with dresses and denim categories.

Native AI Shopping Experiences

Major brands are building proprietary AI tools alongside third-party integrations:

• Home Depot: Magic Apron for intelligent on-site support
• Ralph Lauren: Ask Ralph for personalized outfit recommendations
• Microsoft: New AI-powered shopping experience built on Stripe

Key Challenges to Watch

1. Security and Trust

How do you ensure an agent stays within its permissions? Shared Payment Tokens help, but building consumer trust takes time. Complete audit trails and instant cancellation capabilities are essential.

2. Preserving Customer Relationships

When an AI agent buys instead of the customer directly, how does a brand maintain loyalty? This is why companies like Ralph Lauren are launching their own branded AI agents.

3. Data Standardization

Merchants need structured, standardized product data for AI agents to understand and interact with. This means unifying language, attributes, and taxonomy across the entire catalog.

Practical Steps to Get Started

If you run an e-commerce business or retail operation in the MENA region, here is a realistic roadmap:

Phase 1 (Within 3 months):

• Audit your product data: Is it structured enough for an AI agent to parse?
• Start with your best-selling product categories
• Implement clean, well-documented APIs

Phase 2 (3-6 months):

• Integrate agentic payment solutions (like Stripe's Agentic Commerce Suite)
• Build a branded AI shopping agent for your store
• Test with a limited customer segment

Phase 3 (6-12 months):

• Expand to your full product catalog
• Develop automated negotiation and dynamic pricing capabilities
• Analyze agent behavior data to optimize your offerings

What Comes Next

Agentic commerce is not just an upgrade to the shopping experience — it is a fundamental redefinition of the buyer-seller relationship. Within 18 to 24 months, it is expected to move from experimentation to mainstream adoption.

Businesses that start preparing their infrastructure now will hold a strong competitive position. Those who wait will find themselves catching up to a radical transformation in how e-commerce works.

The question is not "Will agentic commerce arrive?" — it is "Are you ready for it?"