The Agentic Commerce Frontier 📅 | May 19 - May 25 2026
Thanks for reading this week’s Agentic Commerce Frontier. This was a pretty big week for Google, as they put a cross-merchant Universal Cart, UCP checkout, AP2 guardrails, AI Mode ads, and merchant-facing AI tooling into the same operating system. We talk about that in this week’s Agentic Commerce Primer.
In other news in the emerging stack beneath “AI shopping”: carts that persist across surfaces, authorization mandates, merchant-of-record boundaries, fraud controls, and new commercial roles built around agent-mediated demand.
Pretty exciting week! Thanks for reading again.
🔥 TL;DR
Google’s Universal Cart is less a shopping feature than a control-plane bid for agentic commerce. By spanning Search, Gemini, YouTube, Gmail, Google Pay, UCP, and AP2, Google is trying to make the cart the persistent workspace where AI-mediated intent becomes a purchasable order.
My takeaway: I was surprised to see so little coverage on this over the past few days . This is a significant play by Google, Read my primer below
Shopify’s Agentic Storefronts terms quietly turn AI channels into formal distribution infrastructure.The terms preserve merchants as seller of record while letting participating channels such as Google and Microsoft surface eligible products through AI-mediated discovery and purchase flows.
My takeaway: Shopify is defending the merchant relationship by making itself the contractual switchboard for agentic distribution; marketplaces should read that as a warning shot
Exa’s $250 million raise at a $2.2 billion valuation makes agent search infrastructure impossible to ignore.Agentic commerce depends on machines finding products, prices, reviews, specs, policies, and availability across a messy web; Exa is betting that LLM-native retrieval becomes a core primitive.
My takeaway: A massive round and very timely to ask the question: Will the critical retrieval/data layer remain an independent horizontal service that many agents use, or will Google bundle that layer into their stack?
Visa’s Agentic Ready programme and the payments-sector debate around consent, authentication, and liability show that the card networks want to shape the rules before agents scale. Tokenization, identity, risk controls, and authorization records are moving from back-office plumbing to strategic product surface.
My takeaway: Those who treat agentic commerce as another ecommerce authorization flow will get commoditized
Pattern Intelligence, Triple Whale’s Moby 2, Zoovu/XGEN AI, and Bloomreach point to a merchant-operator arms race. The market is filling with tools that do not just analyze ecommerce performance but recommend, automate, or execute pricing, merchandising, discovery, and campaign actions.
My takeaway: “Copilot for merchants” is already stale positioning. What’s important is which vendors will be allowed to touch margin, inventory, and paid media budgets without a human in every loop
🤖 Agentic Commerce Primer: Google Was Supposed to Die. Instead, It Moved Down the Stack, And Is More Relevant Than Ever
TL;DR:
Google was supposed to be killed by AI assistants; instead it’s trying to become the infrastructure they all run on. By wiring UCP (a shared language for offers and policies), AP2 (a protocol for what agents are allowed to do with your money), and Universal Cart (a cross‑surface shopping state) into a live stack, Google is shifting from monetizing blue links to monetizing the rails, shortlists, merchant services, and cloud underneath agentic commerce.
Back in 2023, one of the clearest stories in tech was that AI would kill Google Search. Paul Buchheit, the creator of Gmail, said Google could be “only a year or two away from total disruption” and that AI would eliminate the search results page where Google made most of its money. Mainstream outlets picked it up and ran headlines about ChatGPT‑style systems making traditional search engines obsolete within a couple of years.
If an assistant can answer the question, plan the task, and recommend the product, the old logic of ten blue links starts to look like a relic. For much of 2024 and 2025, most commentary assumed the future belonged to whoever owned the query, and Google’s moat looked dangerously tied to an interface AI seemed designed to bypass.
This week’s Google announcements suggest a different story. Google may not need to preserve the old search experience to preserve power. It may only need to make itself hard to avoid further down the stack: in the protocols, permissions, merchant semantics, and transaction state that agent‑driven commerce will rely on, no matter where the first prompt is typed.
The strategic point is not a few new features, but that components introduced over the last nine months now look like a coherent operating layer for agentic commerce.
The obvious story is too small
At the surface level, the news is straightforward.
Google introduced Universal Cart, a persistent shopping hub that lets users add products from Search, Gemini, YouTube, and Gmail into one place, track prices and stock, and move toward checkout in a single flow
It expanded UCP‑powered checkout beyond the United States into Canada and Australia, with the United Kingdom to follow, and into categories like hotels and local food delivery
It continued to evolve AP2, the Agent Payments Protocol, as the permission and audit layer for purchases made by software agents
Read that quickly and it sounds incremental: better shopping UX, more AI, more Google. The deeper point is in the architecture underneath.
AP2 was announced in September 2025 as an open, payment‑agnostic protocol for agent‑initiated transactions and mandates. UCP was introduced in January 2026 as a common language for agentic commerce across merchants, platforms, and payment providers. Universal Cart is the new, user‑visible layer, but the real shift is that these older rails are now attached to a live, high‑distribution product surface.
Google is trying to define the language agents use
These announcements make more sense if you see them as an attempt to define the language economic agents will use to talk to one another.
UCP standardizes what can be said about products, offers, fulfillment, policies, and post‑purchase flows, so agents can reason over them consistently
AP2 standardizes what can be said about authorization, constraints, mandates, and proof of execution when an agent is allowed to act with a user’s money
Together, UCP and AP2 start to look less like APIs and more like a grammar. UCP defines the semantics of what the merchant side is offering. AP2 defines the semantics of what the user side has permitted. Universal Cart becomes the persistent state object between them, carrying shopping intent across surfaces and time.
Winning a single app is one kind of power. Defining the language that many apps have to speak is another. If serious agentic commerce ends up working in UCP‑style offer semantics and AP2‑style mandate semantics, Google does not need every agent to be “its” agent to matter deeply to the system.
A new kind of history: how intent turns into money
Universal Cart is being marketed as convenience: one cart, many surfaces, fewer interruptions. There is more going on.
Once Universal Cart is wired through UCP and AP2, Google gains a structured history of how human intent resolves into actual payments.
Cart captures the episode: what you add, ignore, defer, swap, or finally buy, and which surface started the journey
UCP gives each option explicit semantics: fulfillment choices, policy constraints, eligibility rules, and merchant conditions
AP2 adds the mandate: spending limits, approval requirements, and authorization boundaries under which the agent acted
Over time, that becomes a very different dataset from classic commerce analytics. It is not just “people who clicked this also bought that.” It is “under these constraints, with these permissions, facing these policies, agents and users consistently chose this path and rejected that one.”
A simple example makes the difference concrete.
Ask an assistant: “Book me a refundable hotel in Lisbon next week, under €250 a night, walking distance from the venue, and don’t use my corporate card unless the rate is non‑refundable.”
On Google’s end‑to‑end rails:
The assistant turns that into a structured constraint set and an AP2 mandate: budget ceiling, date range, “refundable unless I approve otherwise,” allowed cards, and a flag for when to loop you back in
It queries UCP‑enabled merchants and platforms, where each room is described with consistent fields for price, distance, refund rules, fees, and availability
It fills Universal Cart with a shortlist that actually respects those constraints and can explain each option (“3 minutes’ walk, fully refundable, personal card; 1 minute’s walk, non‑refundable, corporate card if you approve”)
When you pick one, the booking is executed within the AP2 mandate, and the whole episode (intent, options, policies, decision, payment) is logged in one coherent trail
With agents but without that stack:
The same assistant scrapes or calls a patchwork of hotel and OTA APIs, each with its own way of encoding “refundable,” “partial refund,” and “credit only”
There is no shared mandate model, just internal heuristics about “user prefers refundable and roughly €250,” so no one outside the agent can see what it was actually allowed to do
The shortlist may look similar, but it’s built on guesses about text and UI labels rather than a consistent schema for policies
When something goes wrong, every party has different logs and no common language for what was supposed to happen
Both scenarios are “agentic commerce” since an agent planned and executed a transaction. The difference is that on Google’s rails it lives inside a shared economic language, while everywhere else it is still held together with duct tape and hope.
That is the gap Google is trying to close: move from improvised, agent‑by‑agent plumbing to a consistent, machine‑readable story of how intent turns into money.
Escaping the “death of search” trap
The original “Google is going to die” story said: if assistants replace search boxes, Google loses its choke point. That is roughly true if all of Google’s power is tied to the results page.
These announcements point to a different defense.
Google seems to be assuming that the front door to the web will fragment. Some requests will start in ChatGPT‑style assistants. Some will start in Gemini. Others will start in vertical agents, messaging apps, or business software. In that world, the winning move is not necessarily to own every first query. It is to own enough of the rails under the transaction that other agents still have to pass through your language, schemas, and audit models when money moves.
The real threat was not that human search habits would change. They always do. The real threat was that Google would stop mediating the valuable steps after intent: comparing options, encoding policies, getting authorization, and settling payments. What this week shows is an attempt to relocate that mediation layer from “ranking pages” to “structuring agent‑led transactions.”
The governance angle
Once agents start spending real money, regulators, payment networks, merchants, and risk teams will all ask the same basic questions:
What did the user actually consent to?
Did the agent stay within that mandate?
What did the merchant actually offer?
How do we handle disputes when an agent misinterprets a policy?
AP2 and UCP already contain early answers.
AP2 formalizes mandates, permissions, and proof of execution
UCP formalizes offers, constraints, and merchant policies in a way machines can read
As soon as those concepts show up in live systems, they become templates. Busy regulators and networks like working examples. If AP2‑style mandates and UCP‑style offer semantics start to be treated as the “normal” way to do responsible agentic commerce, the market will quietly bend around them.
That is how this stops being just product news. If this stack becomes the default pattern for “safe agent payments,” Google’s ideas about what a mandate and an offer look like will influence more than its own products.
Power may still sit upstream
None of this means all the leverage has shifted downstream into rails. There is a separate, important thesis: the deepest power in agentic commerce may sit before discovery, in persistent memory and constraint formation.
That is the layer where long‑term context, habits, trade‑offs, and current circumstances become the constraints that define a shortlist, before any ranking happens. The assistant that holds that context and sets those constraints will shape a large portion of what gets bought.
These announcements do not change that. They make the downstream stack stronger. If a memory‑rich assistant eventually becomes where users store life context and delegate planning, it will still need good rails to compare offers, hold state, get permission, and move money. UCP, AP2, and Universal Cart together now look like a serious candidate for those rails.
The open question is not whether Google has already won the memory‑and‑constraints layer. It has not. The question is whether it can become indispensable to any agent that does.
Can anyone challenge Google’s hegemony?
Google can still be challenged, but not by replaying the old search war. Assistant‑first players like OpenAI or Anthropic can win if they become the main place people express intent and store long‑term context, but they still need merchant semantics and safe payment rails, which pushes them back toward patterns like UCP and AP2 unless they build their own.
Merchant and payments coalitions – Shopify, Stripe, PayPal, Worldline, the card networks – can push back by coordinating on alternative standards so no single consumer platform owns the grammar of agentic commerce.
Platform giants and vertical specialists (Apple, Amazon, Meta, travel and healthcare incumbents) have the assets to build their own stacks in specific domains. The more they do, the more Google’s stack becomes one powerful implementation among several, not an automatic monopoly.
So how big should this be made?
This is not the moment Google “wins” agentic commerce. AP2 may not become universal. UCP may fragment. Universal Cart may end up more important inside Google’s walls than outside them. The upstream fight over persistent memory, trust, and constraint setting is still in progress.
But it is bigger than a normal feature drop. The key change is that Google has moved from talking about an agentic commerce stack to running one in public, with real users, real merchants, and real delegated payments on top of shared protocols. If this works, Google’s next trillion in revenue will come from the rails themselves, ads and offers inside the agent’s shortlist, services to merchants who plug into UCP and Universal Cart, and cloud revenue from every third‑party agent that adopts AP2.
The 2023 story was that AI would kill Google Search. The more interesting 2026 story may be that while everyone stared at the search box, Google quietly moved down the stack and started trying to own the language and rails of the agent economy.
🚀 Major Announcements & Funding News
Google introduced Universal Cart as the consumer-facing layer for agentic shopping: The product is designed to work across Google surfaces and retailers, track price and availability in the background, suggest compatible or alternative items, and route checkout through Google Pay or merchant sites while preserving the retailer as merchant of record (blog.google)
Google tied Universal Cart to UCP and AP2 as the transaction-control layer: Google described Universal Checkout Protocol as the checkout path and Agent Payments Protocol as the guardrail system for digital mandates, authorization, and an audit trail when agents act for shoppers (blog.google)
Google Marketing Live extended UCP into retailer and ad products: The company announced select Google Pay checkout across Nike, Sephora, Target, Ulta Beauty, Walmart, Wayfair, and Shopify merchants including Fenty Beauty and Steve Madden, while also pointing to Direct Offers, YouTube Shopping ads, hotel booking, local food delivery, and Affirm/Klarna financing in Google Pay (blog.google)
Google expanded AI Mode ads into the decision layer of shopping: Conversational Discovery ads, Highlighted Answers, AI-powered Shopping ads, Business Agent for Leads, and Direct Offers connect ad inventory with task-oriented customer conversations and purchase execution (blog.google)
Google framed Search itself as an agentic surface: Its I/O Search update described AI features that can reason across queries and perform task-oriented assistance, including agentic behavior that affects discovery, evaluation, and conversion paths (blog.google)
Google’s developer updates pushed agent infrastructure toward mobile and edge contexts: AI Edge Gallery added MCP integration, notifications, and session continuity, while LiteRT-LM and ADK-related updates point toward more local, app-embedded agent behavior for Android and other environments (Google Developers Blog)
Triple Whale launched Moby 2 as an AI ecommerce operator: The release moves Triple Whale beyond analytics into execution, with Moby 2 connected to real-time ecommerce data and designed to recommend and take action across performance, growth, and operating workflows (Triple Whale / PR Newswire)
Pattern launched Pattern Intelligence, or Pi, for autonomous marketplace execution: Pi is positioned as an AI-powered ecommerce execution engine for brands selling across marketplaces such as Amazon, Walmart, TikTok Shop, and eBay, using proprietary data, sensors, and action loops to identify and act on opportunities in real time (Business Wire)
Zoovu acquired XGEN AI to consolidate AI-native product discovery: The deal combines Zoovu’s guided product discovery platform with XGEN AI’s search, merchandising, and recommendation capabilities, creating a broader engine for agent-ready ecommerce discovery (Zoovu / PR Newswire)
Shoplazza won TikTok’s Partner Innovation Award for commerce marketing technology: The award recognizes Shoplazza’s role as an AI-native commerce operating system for merchants using TikTok’s commerce and advertising ecosystem, reinforcing social commerce as an adjacent execution surface for agentic retail workflows (Shoplazza / PR Newswire)
COLIBRIX ONE and BitGN launched a Global Agentic Ecommerce Challenge for developers: The competition connects AI developers with ecommerce businesses seeking to evaluate autonomous commerce agents before production deployment, with the challenge opening May 30, 2026 (Fintech News Switzerland)
Adzymic launched AgenX Creative Agent as the first product in an agentic advertising suite: The release targets interoperable, agent-to-agent creative execution for brands, agencies, and media owners, shifting ad production from manual workflow assembly toward protocol-based creative operations (ExchangeWire)
Exa raised $250 million at a $2.2 billion valuation to build search infrastructure for AI agents: Exa’s Series C, led by Andreessen Horowitz, is material for agentic commerce because product discovery, merchant retrieval, price comparison, and task execution all depend on fresh, high-quality web search for non-human buyers (Exa)
Shopify formalized participating-channel terms for agentic commerce distribution: The May 25 terms incorporate Google Merchant Center and Microsoft Copilot Shopify Checkout terms where enabled, signaling that AI-shopping distribution is becoming a contractual channel-management problem, not just a product-feed problem (Shopify)
And a cheeky last minute addition which I never thought would be a headline: The Pope and Anthropic announce a partnership …
Pope Leo XIV and Anthropic co-founder Chris Olah called for a church-tech ethics partnership on AI: At the Vatican release of Pope Leo XIV’s AI-focused encyclical Magnifica Humanitas on May 25, Olah argued that AI governance needs pressure and guidance from institutions beyond Big Tech, while the pope called for robust regulation and human-centered oversight. The item is best framed as an AI ethics and governance development rather than a commercial Anthropic partnership announcement (Reuters)
🛡️ Security & Fraud
Identity and payments operators flagged AI fraud as a structural constraint on agentic commerce: Secure Technology Alliance’s summit recap highlighted synthetic identities, fake documents, deepfakes, LLM-assisted attacks, and merchant readiness gaps as agent-assisted shopping moves toward more autonomous payments (Secure Technology Alliance)
Visa’s Spring 2026 threat framing showed fraud adaptation on both attack and defense sides: Digital Transactions reported that AI is speeding personalized scams while issuers and networks are using automated detection, triage, and response to reduce fraud outcomes in areas such as token-level attacks and enumeration (Digital Transactions)
Visa’s CEMEA Agentic Ready rollout put identity, tokenization, risk, and transaction controls at the center of agent-led payments: The programme is designed to test how trusted agent-initiated payments can operate before mass consumer adoption, making issuer-side governance a prerequisite rather than a post-launch compliance layer (Arabiya Capital)
Sygnum’s live AI-agent transaction test highlighted custody and key-control boundaries: The bank’s emphasis on client control of private keys points to the same core issue in agentic payments: agents may initiate instructions, but authority, custody, and final control need to remain explicitly attributable. (Fintech News Switzerland)
Bank Underground framed agentic commerce as a payments-infrastructure and regulatory-design problem: The May 21 post focused on identity and authentication across human and agent actors, high-frequency low-value payments, deterministic legal requirements interacting with non-deterministic AI systems, and interoperability between competing standards. (Bank Underground)
Shopify’s May 25 Agentic Storefronts terms clarify seller-of-record and liability boundaries: The terms state that transactions are completed through the merchant’s Shopify Store and that the merchant remains responsible for obligations tied to the customer transaction, an important control point as AI channels intermediate discovery and purchase (Shopify)
Shopify’s terms permit data disclosure to participating channels for transaction servicing, fraud and abuse prevention, and legal compliance: That language matters because agentic commerce expands the number of parties touching product, order, and customer data during discovery, checkout, order-status updates, and post-purchase servicing (Shopify)
📈 Consumer & Market Insights
Google’s AI Mode data suggests assisted shopping is compressing decision time: Google cited Ipsos research saying 75% of people using AI Mode reported faster, more confident decisions, a signal that agentic interfaces may affect not just conversion but the shape and duration of consideration (blog.google)
Merchant visibility is moving into AI answer surfaces: Google said Merchant Center performance insights will help brands understand visibility in AI experiences and will roll out across Australia, Canada, India, New Zealand, and the U.S., making AI-surface share of voice a measurable retail-ops concern (blog.google)
Practical Ecommerce’s May 20 tool roundup showed agentic commerce moving from protocol announcements into merchant operations: The roundup included agentic storefronts, AI sales agents, universal checkout MCP infrastructure, retail media tools, and AI customer-service layers, indicating that the merchant tooling layer is fragmenting quickly (Practical Ecommerce)
Bloomreach’s ecommerce-agent analysis connected AI agents to campaign execution, not just product discovery: The article describes AI agents for ecommerce across recommendations, campaign creation, and cart recovery, including an example where campaign creation time dropped to 15 minutes (Bloomreach)
AI-mediated discovery is becoming a merchant access layer rather than a traffic source: Practical Ecommerce argued on May 25 that agentic commerce changes where customer relationships begin and end, with merchants needing to influence AI agents directly rather than relying only on marketplaces, search, or social algorithms (Practical Ecommerce)
Merchant catalog readiness became more urgent as Shopify’s May 25 terms took effect: Shopify’s Agentic Storefronts framework makes structured product data, eligibility, participating-channel settings, and seller-of-record obligations part of the operating model for AI-channel distribution (Shopify)
🎯 Strategic Hiring Highlights
Product / platform / leadership
Adyen — Commercial Lead, Agentic Commerce — New York, NY; San Francisco, CA — Salary not listed — Adyen Careers
Checkout.com — Product Manager, Agentic Commerce — London, UK — Salary not listed — Checkout.com (Ashby)
Swap — VP/Director of Product, Agentic Commerce — London, UK (hybrid) — Salary not listed — Swap (Ashby)
PayPal — Senior Director, Payment Services Authorization Optimization — San Jose, CA (hybrid) — Salary not listed — PayPal Careers
Mastercard — Director, Agentic Commerce – NAM Product Management — San Francisco, CA (hybrid) — Salary not listed — Mastercard Careers
Accenture — Agentic Commerce Senior Manager | Consumer Goods & Retail — Multiple locations — Salary not listed — Accenture Careers
SAP — AI Product Owner – Agentic Analytics — Prague 5, Czech Republic (hybrid) — Salary not listed — SAP Careers
SAP — Internship: AI Developer – Analytics & Agentic AI — Paris / Île‑de‑France (hybrid) — Paid internship — SAP Careers
Upstart — Principal Product Manager, Agentic Platform — Remote, US — Salary not listed — Upstart Careers
Microsoft — Principal Product Manager, Agentic Experiences — Redmond, WA — Salary not listed — Microsoft AI Jobs
Bread Financial — Senior Product Owner – Agentic AI — Columbus, OH — Salary not listed — Bread Financial Careers
Instacart — Principal Product Manager, Agentic Commerce — Remote (US/Canada) — Salary not listed — Instacart Careers
American Express — Product Director, Agentic Commerce, Strategy and Insights — New York, NY (hybrid) — Salary not listed — American Express Careers
American Express — Senior Product Manager – Agentic Commerce — New York, NY — Salary not listed — American Express Careers
Citi — Head of Agentic Commerce, Commerce Media and Digital Wallets Engineering (Director) — Irving, TX — Salary not listed — Citi Careers
Amazon — Principal Product Manager – Tech, Amazon Catalog – Product Graph — New York, NY — Salary not listed — Amazon Jobs
Bloomreach — Vice President & GM, Product Management, Commerce AI — Location not listed (remote/hybrid per posting) — Salary not listed — Bloomreach (Greenhouse)
GTM / commercial / partnerships / risk
Adyen — Strategic Growth Manager, Agentic Commerce — Amsterdam, Netherlands — Salary not listed — Adyen Careers
Forter — Senior Strategic Partnerships Manager, AI and Agentic Commerce — United States (remote; New York or London preferred) — Salary not listed — Forter (Greenhouse)
Checkout.com — Manager, Product Marketing, AI and Agentic Commerce — London, UK — Salary not listed — Checkout.com (Ashby)
Stripe — Business Development Manager, Agentic Commerce — South San Francisco HQ; Seattle, WA; or Remote US — Salary not listed (role supports Stripe Agentic Commerce Suite GTM) — Stripe Careers
American Express — Manager – Fraud Risk Servicing and Strategy – Agentic Commerce — Phoenix, AZ or New York, NY — Salary not listed — American Express Careers
Swap — Partner Manager, Logistics (Agentic Storefront / Ecommerce OS) — London, UK — Salary not listed — Swap (Greenhouse)
Engineering / data / agent infrastructure
Stripe — Engineering Manager, Agent Experience — Toronto, Canada — CA$213,600–CA$320,400 — Stripe Careers
Stripe — Engineering Manager, Agentic Commerce — New York, NY — $236,000–$354,000 — Stripe Careers
Wizard Commerce — Senior Software Engineer, Agentic Commerce & Automation — Remote, US — Salary not listed — Wizard Commerce (Greenhouse)
Swap — Senior Data Scientist, Agentic AI — London, UK (hybrid) — Salary not listed — Swap (Ashby)
Swap — Agentic Data Alchemist, AI Data Scientist — London, UK (hybrid) — Salary not listed — Swap (Ashby)
Mirakl — AI Agent Builder (Agentic Commerce Infrastructure / Mirakl Nexus) — Boston, MA (hybrid, 4 days office) — $150,000–$190,000 OTE — Mirakl Careers
Robinhood — Senior Machine Learning Engineer, Agentic — Bellevue, WA; Menlo Park, CA — Salary not listed — Robinhood (Greenhouse)
Firework — AI GTM Engineer (commerce video/AI) — Remote, India — Salary not listed — Firework (Lever)
Design / UX and adjacent
Stripe — Staff Product Manager, Support Experience — Seattle, WA; New York, NY; South San Francisco HQ; Remote US — $214,300–$321,500 — Stripe Careers
📖 Articles Worth Reading
Would you let robots spend your money? Google is betting on it: The Verge gives the cleanest read on Universal Cart as a power shift: Google wants the shopper relationship and cart continuity while leaving fulfillment, liability, and merchant-of-record status with retailers (The Verge)
For Now, Agentic Commerce Raises More Questions Than Answers: Digital Transactions focuses on the payment-system questions operators will need to resolve before autonomous or semi-autonomous checkout scales: intent, consent, liability, chargebacks, and AI-specific rules (Digital Transactions)
What Is Agentic Commerce? A Complete Guide for 2026: Gr4vy’s guide is useful for framing the trust shift from merchant-controlled experiences toward agent-mediated selection, prioritization, payment, and dispute flows (GR4VY)
Google’s I/O announcements show how shopping, search, and agentic assistance are being bundled: The Guardian’s roundup places shopping and transaction assistance inside Google’s broader push across Search, Gemini, Android, and AI-enabled consumer interfaces (The Guardian)
All in on AI Series: Agentic Commerce: Microsoft Advertising’s piece frames agentic commerce through advertising, brand discovery, and the shift from human search behavior to AI-mediated demand capture (Microsoft Advertising)
Agentic commerce rewrites payment choice: Accenture’s payments analysis is useful for banks, wallets, and PSPs because it treats agentic commerce as a visibility and trust problem for payment methods, not just a checkout UX trend (Accenture)
Agentic Commerce 101: Preparing Your Catalog for AI Shopping Agents: 1Digital Agency’s May 25 guide is useful at the implementation layer because it translates agentic commerce into catalog structure, product data quality, checkout readiness, and machine-readable merchandising requirements (1Digital Agency)
Shopify Catalog and AI channels: the 2026 guide: Craftshift’s guide breaks down Shopify Catalog, Agentic Storefronts, AI-channel distribution, and the operational consequences of the May 25 terms update for merchants (Craftshift)
🧭 Looking Ahead
Money20/20 Europe
Date: June 2–4, 2026
Location: RAI Amsterdam, Netherlands
Focus: Money20/20 Europe’s 2026 themes include AI operating inside rewired financial ecosystems. This is the payments-side event to watch for tokenization, delegated payments, identity, compliance, and infrastructure that make agentic checkout viable at scale
NRF 2026: Retail’s Big Show Asia Pacific
Date: June 2–4, 2026
Location: Marina Bay Sands, Singapore
Focus: NRF APAC will bring regional retailers and tech providers together around the “connected, AI-native store” and cross-border commerce
Shoptalk Europe 2026
Date: June 9–11, 2026
Location: Fira Gran Via, Barcelona
Focus: Shoptalk Europe’s 2026 agenda already includes sessions on AI readiness, shifting consumer priorities, and retail transformation. For agentic commerce watchers, it should be one of the best places to hear how European retailers are thinking about data readiness, orchestration, and AI-enabled growth
MAG Payments Summit London 26
Date: June 9–10, 2026
Location: Convene, 155 Bishopsgate, London
Focus: European merchant payments, regulation, technology, and operating-model shifts for large commerce organizations
VivaTech 2026
Date: June 17-20, 2026
Location: Paris, France
Focus: European AI, commerce platforms, startups, infrastructure, and enterprise technology partnerships.
CommerceNext Growth Show 2026
Date: June 23-24, 2026
Location: New York, NY
Focus: Ecommerce growth, retail media, digital commerce operations, AI-enabled customer acquisition, and retention.
GITEX AI Europe 2026
Date: June 30-July 1, 2026
Location: Berlin, Germany
Focus: Enterprise AI, cybersecurity, cloud infrastructure, AI governance, and cross-sector technology deployment.
Berkeley Agentic AI Summit 2026
Date: August 1-2, 2026
Location: Berkeley, CA
Focus: Agentic AI research, infrastructure, interoperability, governance, and academic-to-industry transfer
General information only. Not legal, tax, investment, or professional advice. No warranty as to accuracy or completeness. Verify independently and consult your own advisers.
If you believe any information is inaccurate, please contact AgenticCommerce@proton.me and we will make a good-faith effort to review and correct it where appropriate.
Sponsored or affiliate content, if any, is disclosed.