Think about it: An AI agent cruises through your online store, fills a cart with products, and checks out — all without a human touching a keyboard and before you say “Agentic AI”.
Sounds convenient, right? Until you realize you can’t tell if it’s a legitimate shopping assistant or a sophisticated bot about to drain your inventory with fraudulent purchases.
That’s the problem Akamai and Visa are trying to solve. The two companies announced a partnership this week to bring authentication and fraud controls to what they’re calling “agentic commerce”, the emerging world where AI agents shop on behalf of humans.
As AI agents start autonomously browsing and buying things for consumers, merchants face a new identity crisis. They need to answer two questions simultaneously: Is this agent legitimate? And who’s the actual person behind it? Get either wrong, and you’re either blocking legitimate sales or opening the door to fraud.
Building the trust layer
Visa’s answer is the Trusted Agent Protocol, a framework that authenticates AI shopping agents and connects them to the humans they represent. Akamai brings its edge-based behavioral intelligence and bot detection to the mix. Together, they’re building what amounts to a trust layer for AI commerce.
“The promise of agentic commerce hinges on recognition: the fundamental ability to trust an agent acting on someone’s behalf,” says Patrick Sullivan, Akamai’s chief technology officer for security strategy. “We prove both who the agent is and, critically, who it represents.”
The bot surge is here
The numbers explain why this matters now. According to Akamai’s 2025 Digital Fraud and Abuse Report, AI-powered bot traffic has surged 300% over the past year. The commerce sector alone saw more than 25 billion AI bot requests during two months. That’s a massive wave of automated traffic, and merchants need a way to separate the helpful from the harmful.
The Trusted Agent Protocol works by letting AI agents transmit information to merchants showing they’re approved for their specific shopping mission. It provides visibility into the consumer making the transaction and passes payment information through existing checkout flows. Crucially, it’s designed to work with standard web infrastructure. It means the 175 million Visa-accepting merchant locations globally can adopt the protocol without rebuilding their systems.
Why AI leaders should care
For data and AI leaders, this partnership signals three important shifts. First, the infrastructure layer for AI agents is being built right now, and payment networks are playing a central role. If you’re building AI shopping experiences, you’ll need to think about authentication from day one.
Second, the dual-identity problem — verifying both the agent and the human — is becoming table stakes. Your fraud detection systems need to evolve beyond traditional bot detection. An AI agent behaving like a bot isn’t necessarily malicious; it might be exactly what it claims to be.
Third, edge-based intelligence is emerging as a critical infrastructure for agentic commerce. Akamai’s approach analyzes behavior at the network edge before it touches sensitive systems. This gives merchants real-time insight to make trust decisions at scale.
Jack Forestell, Visa’s chief product and strategy officer, puts it simply: “Agentic commerce is unlocking an entirely new wave of digital interactions, but it can only scale if every player in the ecosystem can trust the agents participating in it.”
The race is on to build the rails for agentic commerce. And the companies laying those rails are the ones defining what trust looks like in an AI-powered economy.
Image credit: iStockphoto/JIRAROJ PRADITCHAROENKUL