AI agents can handle your grocery orders pretty well these days. Everything else? Not so much. Despite splashy infrastructure announcements from the biggest names in payments, the reality of AI-powered shopping remains frustratingly narrow.
Stripe Inc. (STNE), Shopify Inc. (SHOP), Visa Inc. (V), Mastercard Inc. (MA), PayPal Holdings Inc. (PYPL) and OpenAI have all rolled out products targeting agentic commerce. The announcements are long on vision and short on specifics, but the underlying technology suggests we're preparing for a future where AI agents execute most consumer transactions. The problem is that the industry is still stuck in phase one, wrestling with fundamental questions about fraud prevention, risk management and how to even identify which agent is doing what.
The Payments Chicken-And-Egg Problem Returns
Payment companies always face the same fundamental question: do you start with merchants or consumers? Issuing banks rarely make good entry points, so the choice usually comes down to one side of the transaction or the other.
Two decades ago, the answer was obvious. Merchant payments dominated as the preferred starting point. Block, Inc. (SQ), back when it was called Square, emerged in 2009 because small merchants were drowning in card processing costs. Local businesses set card payment minimums or went cash-only. Apple Pay didn't exist in the U.S. yet, which created massive demand for merchant-focused payment solutions.
That wave produced some spectacular winners. Stripe built merchant payment technology for startups. Lightspeed Commerce Inc. (LSPD) tackled restaurant payments. Block created an ecosystem for small business processing, banking and capital. Vertical fintech companies like Shopify emerged to serve specific industries. These companies collectively generated trillions in economic value over 25 years.
Why Merchant Acceptance Is Moving So Slowly
In theory, getting merchants to accept AI transactions should be straightforward. Companies want engaged users who actually make purchases and generate revenue. The infrastructure already exists, ready to plug in.
In practice, adoption is crawling along compared to typical tech cycles. Payment industry veterans say this is their fastest-moving project ever, which tells you something about how glacial this industry normally operates.
When Apple Pay launched, large retailers jumped on board quickly while smaller merchants dragged their feet for years. This cycle should be different because platforms like Shopify and PayPal specialize in solving exactly these kinds of technology adoption challenges for merchants.
The real challenge isn't technical capability, it's awareness and motivation. Smaller merchants who've never heard of agentic commerce face serious navigation difficulties. A startup consumer packaged goods brand in Los Angeles operates completely differently from a 40-store furniture retailer in Arizona. Even with the technology available and ready to deploy, someone still has to convince individual merchants to adopt the new protocol. The exception might be the PayPal-Mastercard integration, which according to their public relations materials may bypass some of these adoption hurdles.
The Fraud Problem Nobody Wants To Talk About
Here's where things get interesting. Can AI agents complete purchases without any merchant involvement at all? Tests suggest the answer is yes. The models are smart enough, and computer vision APIs plus computer use capabilities prove reliable enough to navigate existing checkout screens.
So why aren't payment and technology companies deploying this capability more widely? Because agentic fraud concerns are intensifying behind closed doors. This isn't a theoretical problem. Agentic fraud is happening right now.
That's why Stripe and Shopify are building protocol layers instead of just letting agents scrape existing checkout flows. The approach essentially filters transactions, admitting only legitimate purchases initially. But given that autonomous agents already have the technical capability to complete transactions, fraud patterns will inevitably evolve over time. The race is on to stay ahead of the bad actors before AI-powered shopping fraud becomes the next major headache for the payments industry.
