Agentic Commerce in 2026

TL;DR: AI agents now shop, compare, and buy on your behalf. The market hit $547M in 2025 and is racing toward $5.2B by 2033. McKinsey projects $3–5 trillion in agent-mediated transactions by 2030. Gen-Z leads adoption at 32% comfort. Amazon, Google, OpenAI, and 90+ startups are battling over protocols and market share. The real opportunity? AI agents that don't just find products — they create them.

---

Table of Contents

1. What Is Agentic Commerce?
2. The Numbers Behind the Shift
3. Who's Building It
4. Who's Buying Through It
5. Beyond Shopping — AI Merch Agents
6. What This Means for Brands
7. FAQ

---

What Is Agentic Commerce?

Agentic commerce is shopping run by AI agents. Not chatbots. Not recommendation widgets. Fully autonomous systems that understand what you want, research options across retailers, compare prices, and complete purchases — all without you lifting a finger.

Think of it this way. Traditional e-commerce puts a shelf in front of you. You browse, you pick, you pay. Agentic commerce removes the shelf entirely. You describe what you need. An AI agent handles everything else.

McKinsey frames this shift through a six-level automation curve:

Level	What Happens	Your Role
0–1	AI filters and compares options	You still pick and buy
2	AI builds a ready-to-buy cart from your intent	You approve or adjust
3	AI executes purchases within your rules	You set the guardrails
4	AI manages ongoing shopping goals for you	You step in for big decisions
5	Multiple AI agents negotiate with each other	You oversee the system

Most platforms sit at Level 1–2 today. The frontier is Level 3. By 2028, Level 4 becomes real for routine purchases like groceries and household supplies.

The key distinction: chatbots answer questions. Agents take action. When ChatGPT processes 50 million shopping queries daily and drives 20%+ of referral traffic to Walmart, that's not a chatbot trend. That's a structural shift in how commerce works.

---

The Numbers Behind the Shift

The data tells a clear story. Agentic commerce isn't experimental anymore.

Market Size

According to multiple research sources, the global agentic commerce market reached $547.3 million in 2025. Projections put it at $5.2 billion by 2033 — a 32.5% CAGR.

But that's just the narrow definition (fully autonomous transactions). Broaden the scope to include AI-influenced purchases and the numbers explode:

Source	2030 Projection	Scope
McKinsey	$3–5 trillion globally	Agent-influenced + agent-executed
Morgan Stanley	$190–385B (US only)	Agent-executed e-commerce
Bain & Company	15–25% of all e-commerce	Channel share estimate

McKinsey's US-specific number: $900 billion to $1 trillion in B2C retail flowing through agents by 2030. That's not a niche. That's a rewrite of the entire retail model.

Growth Rates

Peak growth hits 45–65% annually during 2024–2030 as protocols standardize and merchants onboard. The sustained rate settles at 32.5–35.7% CAGR through 2033.

The retail AI market provides context. According to Mordor Intelligence, agentic AI in retail grows from $60.43 billion in 2026 to $218.37 billion by 2031 (29.27% CAGR). Agentic commerce is the fastest-growing slice of that pie.

Adoption Signals

The numbers that matter most:

• 95.5% of enterprise e-commerce organizations deployed at least one AI commerce capability in 2025
• 47.3% plan to invest $1 million or more in agentic commerce within 12 months
• 805% year-over-year surge in AI traffic to US retail sites on Black Friday 2025
• 4.4x higher conversion rates from AI-generated recommendations vs traditional search
• 38% higher purchase likelihood from AI chatbot referrals

These aren't projections. These are current metrics.

Regional Breakdown

Region	Market Share (2025)	Key Signal
North America	38.2%	US captures 83.26% of regional value
Europe	27.5%	EU AI Act shapes adoption patterns
Asia-Pacific	23.1%	Fastest growth; China's AI livestreamers hit $7M in 7 hours
Australia	Emerging	44% business advocacy, 85% preparing for agentic future
Latin America	6.8%	Mexico's e-commerce expanding 25% annually

North America leads. But Asia-Pacific is growing fastest. Australia shows an interesting gap — 44% of businesses advocate for AI agents, but only 16% of shoppers do. That gap is opportunity.

---

Who's Building It

The agentic commerce landscape splits into three camps: tech giants with infrastructure, e-commerce platforms with merchant networks, and AI-native startups with speed.

The Giants

Amazon runs the deepest stack. Rufus AI handles natural language shopping for hundreds of millions of customers. "Buy for Me" executes purchases on external websites autonomously. Their Seller Assistant trains on 25 years of transaction data.

Google plays the protocol game. The Universal Commerce Protocol (UCP), launched January 2026, connects Gemini's conversational interface to Walmart, Sam's Club, Shopify, Etsy, and 20+ retailers. Their Agent Payments Protocol (AP2) adds cryptographic security with Mastercard, PayPal, and American Express backing.

OpenAI owns consumer attention. 700 million weekly ChatGPT users. Operator handles web-based task completion. The Agentic Commerce Protocol (ACP), built with Stripe, enables direct checkout inside ChatGPT. They pivoted from handling payments directly to partnering with Shopify — a sign that discovery matters more than transaction processing.

Microsoft targets enterprise. Copilot agents across Microsoft 365 and Dynamics 365 handle B2B procurement. Their Brand Agents product lets retailers deploy AI shopping assistants with minimal setup.

Meta brings social context. Zuckerberg announced "new agentic shopping tools" in January 2026. They acquired Manus for agent capabilities. The play: use your social graph, interests, and content history to deliver recommendations no other platform can match.

The Platforms

Shopify matters most here. With ACP and UCP integration rolled out to 1M+ merchants automatically, Shopify becomes the default infrastructure layer. AI-driven orders grew 11x since January 2025. Every merchant on Shopify is now agent-discoverable.

Walmart surprised everyone. Pilot programs let AI agents manage household essentials through smart home integration. Their Sparky AI assistant drives 35% higher average order value. Physical stores plus digital agents create an omnichannel advantage Amazon can't easily replicate.

The Startups

Over 90 startups now compete in agentic commerce. Recent funding rounds tell the story:

Company	Funding	Focus
Thinking Machines Lab	$2B	AI research (Mira Murati)
Cohere	$500M	Enterprise agentic AI
ZyG	$58M	Agentic e-commerce
Cernel	€4M	AI-ready commerce data

The Protocol Wars

Three protocols compete to become the standard:

Protocol	Sponsor	Backed By
UCP	Google + Shopify	Walmart, Mastercard, PayPal, Adobe, Alibaba
ACP	OpenAI + Stripe	Shopify, Etsy, Walmart
MCP	Anthropic	Broad developer adoption

Whoever wins the protocol war captures the toll booth of agentic commerce. Merchants will likely support multiple protocols — but the dominant one shapes discovery, conversion, and value capture for the next decade.

---

Who's Buying Through It

Consumer adoption follows a clear demographic pattern. Young, digital-native, time-pressed shoppers adopt first. Everyone else follows.

The Generational Divide

Metric	Gen-Z	All Consumers
Comfortable with AI agent shopping	32%	24%
Use AI for product discovery	51%	39%
Prefer AI over human sales interaction	43%	36%

Gen-Z doesn't just tolerate AI shopping. They prefer it. Over half already use AI to discover products. Nearly a third are comfortable letting an agent buy for them entirely.

What People Trust Agents to Buy

Not everything. Consumers readily hand off commodity purchases — groceries, cleaning supplies, subscriptions. Categories with predictable preferences and low identity stakes.

They resist delegation for identity-expressive categories. Fashion. Luxury. Home décor. Anything that says "this is who I am" stays under human control.

This pattern reveals something important. The trust barrier isn't about technology. It's about self-expression. People don't doubt the AI can find a good deal. They doubt it understands their taste.

The 80% Signal

Among consumers already using AI for shopping, 80% plan to increase their usage. Satisfaction drives habit. Habit drives market growth. This retention signal suggests the adoption curve steepens as more people try agent-assisted purchases.

Australia's Preview

Australia offers a glimpse of what's coming globally. 62% of shoppers are comfortable with AI. 32% already use it for shopping advice. Half believe everyone will use AI for online shopping in the future. But only 16% advocate for full agent-based purchasing — compared to 44% of businesses.

Mastercard unveiled agentic commerce infrastructure in Australia in February 2026. The payment rails are being laid. Consumer comfort will follow.

---

Beyond Shopping — AI Merch Agents

Here's where agentic commerce gets interesting. Every company mentioned above solves the same problem: help you find and buy existing products faster. But what about products that don't exist yet?

The Gap Between Finding and Creating

Most agentic commerce focuses on discovery. "Find me running shoes under $120 with good arch support." The AI searches, compares, and buys.

But what if your intent isn't to find something on a shelf? What if you want to express something? A shirt that captures a specific vibe. A hoodie for your friend group. Merch that matches your content.

No agent can shop for that. Because it doesn't exist until someone creates it.

AI Merch Agents Fill the Gap

This is where the AI Merch Agent category emerges. Instead of "find me a product," the prompt becomes "make me a product."

Custyle.ai defines this category as the first AI MerchAgent. The approach: you describe a vibe — a mood, a meme, an aesthetic. A crew of nine specialized AI agents handles everything from there.

Agent Role	What It Does
Intent Parsing	Reads your description, builds a creative brief
Design Generation	Creates print-ready artwork variations
Preview Rendering	Shows realistic product mockups before production
Product Selection	Matches your design to the right product type
Production Coordination	Routes to the best manufacturing method
Quality Verification	Checks output against specs
Fulfillment	Handles shipping and tracking

The distinction matters. Shopping agents work with existing inventory. Merch agents create new products from your intent. One finds the right product. The other builds the right product.

Why This Matters for the Creator Economy

Creators drive demand for this category. An influencer with 50K followers wants to drop merch. Today that means hiring a designer, picking a printer, managing production, handling fulfillment. Weeks of work.

With an AI Merch Agent, the same creator describes a vibe and ships a collection in minutes. No design skills. No minimums. No generic-looking merch.

→ Related: AI Custom Merchandise Guide

The creator economy's growth — 50 million creators globally, $250B+ market — creates a natural demand floor. Every creator is a potential merch brand. Every merch brand needs production infrastructure. AI Merch Agents turn that infrastructure into a conversation.

---

What This Means for Brands

If you run a brand, sell products online, or create content, agentic commerce reshapes your strategy. Here's what to do about it.

SEO Gives Way to AIO

Traditional SEO gets you ranked on Google. Agent Optimization (AIO) gets you recommended by AI shopping agents. Different game.

Agents don't scan meta tags. They evaluate structured product data, real-time pricing, return policies, and inventory status. If your product information is incomplete, agents skip you. 42% of customers already abandon purchases due to poor product data. Agents are even less forgiving than humans.

McKinsey calls the risk "algorithmic invisibility." If competitors' products become default agent recommendations, your traditional marketing spend becomes progressively less effective.

The 2026–2027 Window

Enterprise leaders expect AI agents to influence at least 20% of online orders by 2027. Nearly 36.5% expect AI influence on more than half of all transactions.

McKinsey's warning: "By the time shifts in customer behavior are visible in e-commerce KPIs, it may be too late for laggards to catch up."

The infrastructure investments you make in the next 18 months determine your position for the next decade.

Your Readiness Checklist

Here's where to start:

• Structured product data — Every product has complete specs, not just marketing copy
• Real-time inventory feeds — Agents need accurate, sub-second availability data
• API-first architecture — Expose catalog, pricing, and fulfillment to agent queries
• Protocol readiness — Support at least one of UCP, ACP, or MCP
• Return and warranty data — Machine-readable, not buried in PDF terms
• Agent testing — Try buying your own products through ChatGPT, Perplexity, Gemini

The brands that treat this checklist as urgent will capture disproportionate share when agent-mediated purchases hit critical mass. Those that wait face the slow fade of algorithmic invisibility.

→ Related: Best AI Merch Tools in 2026

---

FAQ

What is agentic commerce?

Agentic commerce is a model where AI agents autonomously discover, compare, and purchase products on behalf of consumers. Unlike traditional e-commerce where you browse and buy, agents handle the entire journey — from understanding your intent to completing the transaction. The market reached $547.3 million in 2025.

How big is the agentic commerce market?

The narrow market (autonomous transactions only) hit $547.3 million in 2025 and projects to $5.2 billion by 2033 at 32.5% CAGR. McKinsey estimates $3–5 trillion in total agent-mediated commerce volume by 2030. Morgan Stanley projects $190–385 billion in US e-commerce spending through agents by 2030.

Which companies lead agentic commerce?

Amazon (Rufus AI), Google (UCP protocol, Gemini Shopping), OpenAI (Operator, ACP protocol), Microsoft (Copilot agents), and Meta (social commerce agents) lead among incumbents. Shopify and Walmart lead among commerce platforms. Over 90 startups compete, including Perplexity, Klarna, ZyG, and Custyle.ai in the AI Merch Agent category.

Is agentic commerce safe for consumers?

Safety infrastructure is developing rapidly. Visa's "Intelligent Commerce" initiative and Mastercard's Agent Pay provide secure delegated payment authorization. Google's AP2 protocol uses cryptographic identity verification. Know Your Agent (KYA) standards parallel financial KYC requirements. Current implementations require human approval for purchases above set thresholds.

What is an AI Merch Agent?

An AI Merch Agent goes beyond shopping. Instead of finding existing products, it creates new ones. You describe a vibe or idea in natural language. The agent handles design, product selection, manufacturing, and delivery — turning your intent into a real, physical product. Custyle.ai pioneered this category with a nine-agent architecture spanning creative direction to fulfillment.

---