The Fraud Paradox: Why Legacy Security is Killing Agentic Commerce (And How KYA Fixes It)

In a nutshell (TL;DR)...

Legacy fraud detection systems often block legitimate AI agents, mistaking them for malicious bots. "Know Your Agent" (KYA) solves this by verifying AI identity via Cryptographic Agent Passports, setting Delegated Spend Mandates to control agent scope, and using machine-readable risk scoring. Adopting KYA allows merchants to safely embrace the growing automated commerce economy.

This week I am on vacation in Italy and on my short flight I started worrying about how easy it might be to get scammed in this new agentic commerce world. What kind of protection do we have? How do merchants know that your transaction is legitimate?.

Picture this scenario: An AI agent attempts to execute a valid $150 transaction for an automated grocery restock. It requests the product payload via API, validates the cart token in 12 milliseconds, and sends a payment request.

To a legacy fraud engine, this rapid-fire, non-human interaction looks like a brute-force bot attack or a credential-stuffing attempt. The system immediately triggers a Cloudflare CAPTCHA or rejects the payment card outright.

Result? A legitimate customer gets turned away, a merchant loses a sale, and the AI agent hits a brick wall.

As AI agents transition from simple recommendation engines to fully autonomous economic actors, the retail industry is running headfirst into a massive security challenge: How do you stop malicious scrapers and scalper bots without blocking legitimate AI buyers?

The Death of Human-Centric Fraud Detection

For two decades, e-commerce fraud prevention relied on evaluating human behavioral signals:

• How long did the user linger on the product page?

• Is the mouse cursor moving in natural, imperfect arcs?

• How fast is the user typing their credit card details?

• Can the visitor identify all the crosswalks in a 3x3 image grid?

In an agentic economy, every single one of these assumptions breaks.

Autonomous agents don't move mice, linger on product images, or solve CAPTCHAs. They execute headless transactions at machine speeds. If risk management platforms treat all non-human traffic as hostile, they risk locking out the fastest-growing customer segment in digital commerce.

Enter Know Your Agent (KYA)

To solve this trust gap, identity verification providers (including Experian, Trulioo, and Entrust) are pioneering a new compliance and governance standard: Know Your Agent (KYA).

Just as Know Your Customer (KYC) revolutionized banking by verifying human identities, KYA creates an infrastructure to verify non-human actors and establish their operational authority.

        THE KYA TRUST TRIAD
               
      +-------------------+
      |   VERIFIED HUMAN  |
      |  (Account Owner)  |
      +-------------------+
                |
  Delegated     |     Cryptographic
  Mandate       |     Binding
                v
      +-------------------+
      |  AUTHENTICATED AI |
      |   (Digital Agent) |
      +-------------------+
                |
  Authorized    |     Scoped
  Intent        |     Limits
                v
      +-------------------+
      |  TRANSACTION DATA |
      | (Cart & Merchant) |
      +-------------------+

Instead of evaluating how a page was navigated, a KYA-compliant merchant system evaluates the Trust Triad:

1. The Human (Identity): Is this agent bound to a real, verified individual or organization?
2. The Agent (Passport): Does the agent present a valid, tamper-proof credential (a Digital Agent Passport) issued by a trusted entity?
3. The Intent (Authorization): Has the human owner explicitly granted this agent a delegated mandate to spend up to a specific dollar amount for a specific category?

3 Core Components of the KYA Stack

When an AI agent checks out at a modern merchant platform, three security mechanics validate the order behind the scenes:

1. Cryptographic Agent Passports

Instead of exposing raw API keys or static credentials, agents carry a Digital Agent Passport (DAP) or cryptographic token signed by an identity registry. When requesting a checkout endpoint, the agent presents this token, immediately proving its publisher (e.g., OpenAI, Google, Anthropic) and its active verification status.

2. Delegated Spend Mandates

To limit blast radius if an agent is compromised or subjected to prompt injection attacks, KYA enforces strict scope boundaries. A user might grant an agent a tokenized spend mandate: "You are authorized to spend up to $200 on running shoes before midnight on Friday." If the agent attempts to purchase a $1,000 television, the transaction fails at the gateway level regardless of payment card limits.

3. Machine-Readable Risk Scoring

Rather than looking for device fingerprints or browser headers, next-gen fraud models evaluate machine-native signals:

• Has this agent's digital passport been revoked?
• Is the merchant endpoint receiving requests consistent with the agent's stated policy parameters?
• Is the transaction origin signed by a verified enclave or zero-knowledge proof?

The Strategic Choice for Merchants

E-commerce brands face a clear strategic fork in the road:

• Path A (The Defensiveness Trap): Double down on legacy bot detection, block headless browsers, enforce aggressive CAPTCHAs, and inadvertently shut out millions of dollars in automated customer orders.
• Path B (The KYA Highway): Implement agent-friendly APIs with clear authentication endpoints, accept cryptographic agent passports, and capture market share in an increasingly automated retail landscape.

The future of digital commerce isn't about choosing between security and automation. It's about establishing a verified layer of identity where humans, agents, and merchants can trade with absolute trust.

The Death of Brand Loyalty: How to Win When Machines Do the Shopping

In a nutshell (TL;DR)...

As AI buyer agents increasingly handle consumer purchasing, traditional brand loyalty built on marketing and visuals is being replaced by "operational excellence." To succeed, brands must move from emotional storytelling to machine-readable data (GXO), focusing on transparent product attributes, API-integrated loyalty programs, and superior post-purchase reliability. Future growth depends on seamlessly fitting into the decision logic of the machines making the purchase.

I decided to dive back into the world of Agentic Commerce this week and it made me wonder how this new concept might alter the way in which we sell, or are sold products. For the last twenty years, digital brand loyalty was built on visual friction and psychological hooks. E-commerce teams spent billions optimizing digital storefronts: high-resolution photography, emotional storytelling, banner ads, exit-intent popups, and strategically placed "You Might Also Like" recommendations.

Now imagine a world where your end customer never visits your website, never sees your homepage, and never interacts with your marketing copy.

As agentic commerce accelerates, consumer purchasing decisions are increasingly delegated to AI models embedded in tools like ChatGPT, Google Gemini, or native shopping super-agents. These AI buyer agents aren't swayed by slick web design, brand nostalgia, or clever marketing copy. They evaluate structured parameters: price, delivery promises, return flexibility, and micro-review sentiment.

When the storefront disappears, traditional brand loyalty dissolves with it. Here is how the retail landscape is shifting and how forward-looking brands are rebuilding loyalty for an AI-first world.

1. The Cold Rationality of the Machine Buyer

When a human shops for coffee beans, they might buy a specific brand because of a colorful bag design, an inspiring origin story on the landing page, or habit.

When an AI agent shops for coffee beans, it executes a parameter-based search:

• Medium roast, single-origin, roasted within 7 days.

• Delivered to doorstep by Thursday at 9 AM.

• Maximum price: $18.00 per 12oz bag.

• Sentiment threshold: Minimum 90% positive sentiment across recent reviews regarding freshness.

If a brand fails any single parameter, the agent bypasses it instantly. No second chances, no emotional leeway. This shifts market dynamics from emotional affinity to operational excellence.

2. Fighting Back with Native "Super-Agents"

Retail giants aren't surrendering the customer relationship to third-party chatbots without a fight. Rather than watching consumers shop through external tools, major retailers are deploying their own AI super-agents natively inside their ecosystems.

A prime example is Walmart’s Sparky, an agentic shopping assistant designed to transform search into a goal-driven, conversational journey.

 TRADITIONAL JOURNEY           AGENTIC RETAIL JOURNEY
 
  [ Keyword Search ]            [ Goal Statement ]
          │                            │
          ▼                            │    

 [ Scroll 50 Listings ]       [ Native AI Super-Agent ]
          │                            │
          ▼                   (Sparky / Native Agent)
[ Compare Prices/Specs ]                │
          │                  ┌─────────┴─────────┐
          ▼                  ▼                   ▼
  [ Manual Checkout ]    [ Multi-Step      [ Cart & Instant
                          Planning ]          Execution ]

Instead of forcing shoppers to type keywords and scroll through hundreds of sponsored items, Sparky handles multi-step goals such as "Plan a backyard BBQ for 10 people under $150 with gluten-free options" and constructs a complete, execution-ready cart directly within Walmart's ecosystem.

By deploying proprietary brand agents, retailers keep the consumer inside their own branded touchpoints while offering the effortless speed of AI automation.

3. Post-Purchase Experience is the New Brand Loyalty

When AI agents evaluate products, past purchase history and customer satisfaction signals heavily influence future recommendations. This means loyalty is no longer won at discovery; it is earned post-purchase.

If an AI agent orders a product for a user and the order arrives late, damaged, or creates return friction, the agent's memory updates. The next time the user asks for a similar item, the agent deprioritizes that brand.

            THE POST-PURCHASE LOYALTY LOOP
               
      +----------------------------------+
      |      AI Agent Executes Order     |
      +----------------------------------+
                      │
                      ▼
      +----------------------------------+
      |  Post-Purchase Delivery & Support|
      +----------------------------------+
                      │
        ┌─────────────┴─────────────┐ 

         ▼                           ▼
[ Smooth Experience ]       [ High Friction ]
        │                           │
        ▼                           ▼

Agent Flags Brand as      Agent Deprioritizes Brand 

   High-Trust Preference        in Future Queries

Brands winning in the agentic era are shifting focus to post-purchase automation:

• Instant WhatsApp/SMS Support: Automated agents resolving exchanges or tracking updates in real time.

• Hassle-Free Returns: One-click return labels generated directly via machine-readable APIs.

• Proactive Replenishment: Post-purchase agents tracking usage cycles and prompting automated re-orders right before a product runs out.

4. Micro-Reviews & Machine-Readable Trust

In the human-centric web, a product with a 4.8-star rating and 10,000 generic reviews like "Great product!" performed well.

AI agents look deeper. They analyze micro-reviews—extracting specific entity sentiment to verify if a product meets granular, real-world constraints:

• "Runs small in the shoulders."

• "Battery lasts 8 hours on high brightness."

• "Waterproofing holds up in heavy rain."

If an agent is tasked with finding a jacket for high-altitude trekking in wet conditions, it reads these unstructured review fragments to calculate a confidence score. Brands that encourage detailed, attribute-rich customer feedback give AI agents the empirical proof needed to select their products over a competitor's.

The New Playbook for Brands

To remain resilient as shopping agents take over the discovery funnel, brands must adopt three fundamental rules:

1. Optimize for Machine Readability (GXO): Move from standard SEO to Generative Experience Optimization (GXO). Ensure product attributes, stock status, delivery promises, and margin rules are accessible via real-time APIs.

2. Elevate Operational Integrity: In an agentic economy, a missed delivery window or an out-of-stock cancellation is not just a lost order, it damages your score in the agent's decision logic.

3. API-ify Your Loyalty Program: Make your reward tiers, points balances, and exclusive perks queryable by external agents so machine buyers can factor member discounts into checkout decisions.

The brands that thrive won't be those with the loudest ad campaigns; they will be the ones that seamlessly fit into the decision logic of the software making the purchase.

The HTTP of AI Shopping: Inside UCP, MCP, and the Open Protocol Stack

In a nutshell (TL;DR)...

This document explores the shift toward "Agentic Commerce," where AI agents independently execute shopping tasks. This ecosystem is powered by three key open protocols: MCP for live data access, UCP for transaction management (carts, identity, and payments), and A2A for dynamic negotiations between software agents. These open standards are vital for a friction-free, competitive future in retail.

In a previous post I talked about the rise of Agentic Commerce and how it will transform the way we shop online. I asked you to imagine asking your Personal AI Assistant: "Find me a waterproof hiking jacket in olive green under $200, apply my store rewards, and ship it to my apartment by Friday."

Five years ago, that prompt would yield a list of web links. Today, the AI Agent doesn't just find the jacket, it checks real-time inventory, verifies your loyalty tier, negotiates dynamic discounts, and executes the purchase without you ever loading a checkout page.

I’ve also delved into the magic sauce that makes this happen but let’s try and bring together the main components into one post and try to understand what is at play. It is an underlying stack of open protocols, a collection of open-source technical standards forming what experts are calling the "HTTP of Agentic Commerce".

If you want to understand how software will trade billions of dollars over the next decade, you need to understand the protocols making it happen: MCP, UCP, and A2A.

1. The Context Layer: Model Context Protocol (MCP)

Before an AI agent can buy anything, it needs access to real-time information. Language models are inherently isolated; they don't natively know if a store has 3 items left in stock or if a price dropped five minutes ago.

Enter Model Context Protocol (MCP), open-sourced by Anthropic.

MCP acts as a universal adapter between AI models and external data tools. Think of it as USB-C for AI: instead of developers writing custom API connectors for every single product database, store backend, or CRM, MCP provides a standardized format for agents to query live data.

What MCP Handles in E-Commerce:

• Live Inventory Checks: Quoting real-time stock counts across multiple warehouse locations.

• Spec Parsing: Extracting structured technical specifications (e.g., precise dimensions, fabric weight, voltage) from unstructured databases.

• Contextual Inputs: Feeding user preferences, sizing profiles, and past purchase histories securely to the AI model.

2. The Commerce Layer: Universal Commerce Protocol (UCP)

While MCP provides data context, it isn't built to orchestrate end-to-end retail transactions. Querying an API for stock is easy; creating a multi-item cart, applying promotional codes, initiating identity verification, and managing payments requires a formal commerce standard.

To solve this, Google teamed up with retail and infrastructure giants—including Shopify, Etsy, Target, Walmart, Visa, and Stripe—to release the Universal Commerce Protocol (UCP).

+----------------------------------------------------+
|               CONSUMER AI SURFACES                |
|     (Gemini, AI Search, Custom Assistants)         |
+----------------------------------------------------+
                            |
                            v
+----------------------------------------------------+
|        UNIVERSAL COMMERCE PROTOCOL (UCP)           |
| Standard primitives for Cart, Identity & Checkout |
+----------------------------------------------------+
            /              |              \
            v               v               v
  +---------------+---------------+--------------+
  |    MERCHANT  |   LOYALTY &   |    PAYMENT  |
  |    BACKEND   |    IDENTITY   |   PROVIDERS |
  +---------------+---------------+--------------+

UCP defines an open-source communication specification for e-commerce operations. It doesn't replace existing merchant platforms like Shopify or WooCommerce; it gives AI agents a set of standard functional primitives to talk directly to them.

Core UCP Primitives:

1. Catalog & Discovery: Exposing structured, real-time product catalogs directly to AI crawlers.

2. Cart Management: Enabling agents to create, modify, and calculate sub-totals for multi-item carts programmatically.

3. Identity & Loyalty Linking: Recognizing that "John Doe" is a Gold Tier member at a store, applying member pricing automatically without forcing a manual login on a website.

4. Checkout & Orchestration: Tokenizing payments and completing transactions securely while keeping the merchant as the official Merchant of Record.

3. The Negotiation Layer: Agent-to-Agent Protocol (A2A)

The ultimate evolution of agentic commerce isn't just a buyer bot interacting with a static website backend—it's software negotiating with software.

Agent-to-Agent (A2A) protocols establish rules for buyer agents (representing consumers) and seller agents (representing brands) to interact dynamically.

How A2A Dynamics Work in Practice:

• Dynamic Bundling: A buyer agent requests a camera, lens, and tripod. The brand's seller agent calculates real-time margins and responds: "If you buy all three together, I can offer an instant $45 bundle discount."

• Inventory Clearing: A merchant's AI agent notices excess seasonal inventory and dynamically grants targeted discounts to consumer agents searching for deals in that specific category.

• Automated Terms Negotiation: Negotiating bulk order delivery timelines or return window extensions for enterprise purchases.

The Big Shift: Why Open Protocols Win

Why does this open architectural stack matter so much? Because walled gardens create friction, and friction kills conversion.

If every AI assistant required a proprietary integration to buy from every merchant, only massive platforms (like Amazon) would survive. Open standards like UCP, MCP, and A2A democratize the landscape. They allow a boutique clothing store running on a standard e-commerce platform to sell to a consumer using ChatGPT, Gemini, or a standalone personal AI assistant seamlessly.

The web was built on open transport protocols like HTTP and HTML. The era of agentic commerce is being built on data, transaction, and negotiation protocols. Brands that adopt these standards early won't just keep up, they will be the first ones discovered when software starts doing the shopping.

Agentic Commerce: The Virtual Procurement Revolution

In a nutshel (TL;DR)...

Agentic AI is transforming B2B procurement by automating routine sourcing and negotiation, shifting operations from reactive to predictive purchasing, and leveraging structured data to allow human experts to focus on strategic oversight through a "Human-in-the-Loop" approach, ultimately reducing operational friction.

The B2B Revolution: Meet Your New Virtual Procurement Department

Earlier this year I posted about the rise of Agentic Commerce being the future of online shopping taking over the consumer experience. We imagined a digital assistant booking a flight to Barcelona or effortlessly picking out the perfect pair of trail running shoes for a weekend getaway. It is an exciting vision, but while consumer applications get most of the spotlight, a much quieter, and arguably much more impactful, revolution is taking place in the business-to-business (B2B) sector.

Welcome to the era of Agentic Commerce in B2B, where AI is moving from a simple data-crunching tool to an active, independent participant in your supply chain.

For years, B2B purchasing has involved a lot of heavy lifting: tracking down supplier catalogs, manually comparing specifications across sprawling spreadsheets, managing endless email chains for quotes, and reacting to sudden inventory shortages. Today, Agentic AI is stepping in to change that dynamic entirely, acting as a "Virtual Procurement Department" for trading and manufacturing companies.

From now on, when I talk about “AI Agents” I am referring to quite a number of software vendors that are offering this as a solution such as ORO Labs or LevelPath that have built their platforms from the ground up with artificial intelligence as the foundation, to the likes of Tropic that lean into AI for highly targeted areas of procurement, such as specialized software (SaaS) purchasing, contract intelligence, or complex e-auctions 

So let’s take a closer look at how platforms like this are transforming B2B commerce from a tedious chore into a highly predictive, streamlined operation.

Say Hello to Your Autonomous Sourcing Analyst

In the traditional B2B model, sourcing new components or managing supplier requests requires a significant amount of manual labor. Procurement teams spend hours sending out Requests for Quotations (RFQs), waiting for responses, and then lining up the data to make a choice.

With the introduction of Agentic Commerce, AI agents can take over this routine work. Instead of merely organizing data, these agents act like intelligent analysts. When a business needs a component, the agent can autonomously search through supplier databases, gather bids, and compare terms. It doesn't just look at the bottom-line price, either. The agent can evaluate lead times, review the supplier's quality history, and select the most cost-effective solution. It can even negotiate discounts within predefined rules and automatically approve orders that fall under specific budget limits.

Moving from Reactive to Predictive Purchasing

If you have ever had to deal with the headache of a sudden supply shortage, you know that reactive purchasing is stressful. Traditionally, businesses order new stock based on historical sales data and a little bit of guesswork, which often leads to either costly overstocking or frustrating "out of stock" scenarios.

Agentic AI shifts the paradigm from reactive to predictive. By analyzing historical data, current sales velocity, and market trends, AI agents can accurately forecast future demand. Furthermore, these agents can continuously monitor inventory levels and automatically trigger restocking orders at the perfect moment to prevent supply chain bottlenecks. The system knows not just what you need to order, but exactly when and in what quantity, ensuring operations run smoothly in the background.

Taming the Messy Supplier Data

Of course, for an AI agent to make smart purchasing decisions or negotiate with suppliers, it needs perfectly structured and accurate data. In the B2B world, supplier data is notoriously messy, often arriving in unstructured formats like massive PDFs or inconsistent spreadsheets.

To support an AI-based procurement platform, a Product Information Management (PIM) system acts as the critical intelligence hub. For an AI procurement agent to accurately analyze supplier bids, forecast demand, and execute purchases autonomously, it must base its decisions on flawless, highly structured data. They can even detect inconsistencies or missing attributes and populate them into the right fields. Ultimately, by guaranteeing that supplier records are accurate, consistent, and complete at the point of ingestion, an PIM system provides the trusted data foundation necessary for a B2B procurement platform to operate autonomously and reliably 

What Happens to the Human Experts?

With agents sending RFQs, negotiating prices, and monitoring inventory, it is natural to wonder where human professionals fit into this new landscape. The good news is that the rise of agentic AI is not about eliminating the procurement team; it is about elevating them.

This brings us to the concept of the "Human-in-the-Loop" (HITL). While AI agents are fantastic at analyzing data and executing routine purchasing tasks, they lack the nuanced judgment, ethical grounding, and relationship-building skills required for high-stakes decisions. In an agentic B2B environment, human experts transition from tactical execution (like manual data entry) to strategic oversight.

For example, an agent might handle 90% of routine supplier restocking autonomously. But if it encounters an unprecedented supply chain disruption, a wildly out-of-budget price hike, or a scenario that requires complex ethical reasoning, it is programmed to pause and escalate the issue to a human manager. Humans remain in control, defining the agent's objectives, setting the financial guardrails, and managing the exceptions.

A Smarter Way to Do Business

The B2B revolution powered by Agentic Commerce is ultimately about removing friction. By handling the drudgery of data processing, supplier negotiations, and inventory tracking, AI agents free up human professionals to focus on what they do best: building strategic partnerships, exploring new market opportunities, and driving innovation.

The future of B2B is predictive, highly automated, and incredibly efficient. And with a Virtual Procurement Department working tirelessly in the background, businesses can look forward to a much smoother ride.