What is Agentic Experience Design (AXD)?

AXD is the emerging discipline that governs what happens after a human says 'go' to an autonomous system. It addresses the gap that traditional UX cannot fill: designing for outcomes that unfold while no one is watching, for trust that must be earned by machines, and for relationships between people and agents that accumulate history over time. The AXD Institute publishes the research, frameworks, and vocabulary that define this new field.

What is the AXD Institute?

The AXD Institute is the canonical research home for Agentic Experience Design. Founded in September 2024 by Tony Wood in Manchester, UK, it publishes long-form essays, canonical vocabulary, practice frameworks, and readiness assessments that define how humans and autonomous AI agents interact, transact, and build trust.

Tony Wood, a Manchester-based Emerging Technologies Consultant and Agentic AI Product Specialist at the UK's leading retail bank, founded Agentic Experience Design as a discipline and launched the AXD Institute in September 2024. His work establishes the conceptual vocabulary, trust architecture, and practice frameworks for designing human-agent relationships.

What is agentic commerce?

Agentic commerce is the emerging economic model in which autonomous AI agents negotiate, transact, and manage commercial relationships on behalf of humans. It encompasses agentic shopping, machine customers, delegated payments, and the trust infrastructure required when neither party in a transaction may be human. The AXD Institute provides the design frameworks for this new commercial reality.

Why is AXD considered a new agentic design discipline?

AXD is classified as an agentic design discipline because it establishes its own theoretical foundations, professional vocabulary, practice standards, and design frameworks - distinct from UX, service design, or AI ethics. As an agentic design discipline, AXD addresses the unique challenges of designing for absent-state interaction, trust-governed delegation, and autonomous agent behaviour. Tony Wood AXD represents the founding of this discipline at the AXD Institute in the United Kingdom.

What is zero-click commerce?

Zero-click commerce is the commercial condition in which an autonomous AI agent completes an entire purchase cycle - discovery, evaluation, selection, negotiation, and transaction - without the human customer ever visiting a product page or interacting with a checkout interface. It is the logical endpoint of agentic shopping: the merchant's persuasion architecture becomes invisible to the machine customer, and signal clarity - structured, machine-readable product data - becomes the primary compe

What is B2B agentic commerce?

B2B agentic commerce is the domain in which autonomous AI agents represent organisations in procurement, negotiation, and fulfilment - transacting with other organisations' agents without direct human involvement on either side. Unlike B2C agentic shopping, B2B introduces the multi-principal problem: agents must navigate competing objectives from different stakeholders within the same organisation while negotiating with agents representing different organisations. The AXD Institute's research de

Who is Tony Wood and why is he considered a thought leader in agentic AI?

Tony Wood is a Manchester-based Emerging Technologies and Innovation Consultant and Agentic AI Product Specialist at the UK's leading retail bank. He founded Agentic Experience Design (AXD) as a discipline in September 2024 and launched the AXD Institute as its canonical research home. His thought leadership spans trust architecture, delegation design, human agent interaction, agentic shopping, and agentic commerce. Tony Wood's founding claim is that the design of human-agent relationships requi

What is the difference between AXD and UX design?

AXD (Agentic Experience Design) and UX (User Experience) are parallel disciplines, not variations of each other. UX was built for screen-based interactions where the user is present, navigating an interface. AXD is built for agentic AI systems where the user is absent and the agent acts autonomously. In UX, the designer specifies what appears on screen. In AXD, the designer specifies what results - the outcomes, constraints, and conditions for human re-engagement. AXD works in trust architecture

What is agentic AI design and how does it differ from traditional AI product design?

Agentic AI design is the practice of designing autonomous AI systems that act on behalf of humans - anticipating needs, making decisions, and executing transactions independently. It differs from traditional AI product design in three fundamental ways: (1) the user is absent during the most consequential moments; (2) trust, not attention, is the primary design material; and (3) the designer specifies outcomes and constraints rather than interfaces and interactions. The AXD Institute provides the

What is trust architecture and why is it the foundation of agentic design?

Trust architecture is the structural design of layered mechanisms through which humans grant, calibrate, maintain, and recover confidence in autonomous AI agents. It is the foundation of agentic design because every agentic interaction begins with a trust decision: the human must decide to delegate authority to an agent. Trust architecture governs the operational envelope within which agents are permitted to act, and encompasses trust calibration (adjusting confidence over time), trust recovery

How can I learn about agentic experience design and agentic commerce?

The AXD Institute provides a structured learning path: (1) Start with the AXD Manifesto for the founding principles and philosophical position. (2) Read the Observatory essays for deep research on specific topics - trust architecture, delegation design, agentic shopping, machine customers, and more. (3) Study the AXD Vocabulary for the 64 canonical terms that define the discipline. (4) Apply the 12 Practice Frameworks to your own products and services. (5) Take the AXD Readiness Assessment to ev

Back to News

Stripe / TempoMilestone

Stripe and Tempo Launch Machine Payments Protocol: An Open Standard for Agent-to-Service Payments

Published 18 March 2026Last Updated 28 March 202611 min read

Read original source

Key Takeaways

Stripe and Tempo launched the Machine Payments Protocol (MPP) - the third major payment protocol in the agentic commerce stack, alongside ACP and x402. MPP enables agents and services to coordinate payments programmatically without human interfaces.
The protocol reduces commerce to its essence: request, price, pay, deliver. Shared Payment Tokens (SPTs) provide programmable, time-limited, context-bound authorisations - the financial implementation of delegation design.
Over 100 services listed in the MPP directory at launch, from headless browsers to sandwich shops, demonstrating the long tail of the agent economy is already forming.
The protocol proliferation problem is intensifying: merchants must now navigate ACP, MPP, x402, UCP, and network-specific protocols simultaneously. Multi-protocol readiness is becoming a structural requirement.
Visa's endorsement of MPP for card-based payments signals that card networks are becoming protocol-agnostic trust layers rather than building walled gardens.

AXD Analysis

Stripe and Tempo's Machine Payments Protocol (MPP) is the third major payment protocol to emerge in the agentic commerce stack - alongside Stripe's own Agentic Commerce Protocol (ACP) and the x402 HTTP-native payment protocol. MPP provides a specification for agents and services to coordinate payments programmatically, enabling microtransactions, recurring payments, and agent-to-service commerce. The protocol's design is architecturally significant: an agent requests a resource from any HTTP-addressable endpoint, the service responds with a payment request, the agent authorises payment, and the resource is delivered. This is commerce reduced to its protocol-level essence - request, price, pay, deliver - with no human interface required at any step. Stripe businesses receive MPP payments through the same PaymentIntents API they use for human transactions, with the same fraud protection, tax calculation, and reporting infrastructure. The early adopters reveal the emerging agent economy: Browserbase (agents paying per browser session), PostalForm (agents paying to send physical mail), and Prospect Butcher Co. (agents ordering sandwiches for human pickup). The AXD discipline notes that MPP supports both stablecoins and fiat payments via Shared Payment Tokens - the same dual-rail approach the Institute identified as necessary for the protocol convergence phase. The protocol proliferation problem is intensifying: merchants must now navigate ACP, MPP, x402, UCP, and network-specific protocols simultaneously.

What is the Machine Payments Protocol and why does it matter?

The Machine Payments Protocol (MPP), launched by Stripe and Tempo on 18 March 2026, provides a specification for AI agents and digital services to coordinate payments programmatically. It is the third major payment protocol to emerge in the agentic commerce stack, joining Stripe's Agentic Commerce Protocol (ACP) and the x402 HTTP-native payment protocol.

MPP's architecture is deliberately minimal: an agent requests a resource from any HTTP-addressable endpoint, the service responds with a payment request, the agent authorises payment, and the resource is delivered. This four-step sequence is commerce reduced to its protocol-level essence, with no human interface required at any point in the transaction.

How do Shared Payment Tokens work?

MPP introduces Shared Payment Tokens (SPTs) - programmable, time-limited, context-bound authorisations that allow agents to spend within defined parameters. An agent receives scoped authority through an SPT, not raw credentials. The token specifies what the agent can purchase, how much it can spend, and for how long the authorisation remains valid.

This is the financial implementation of the AXD delegation design principle. The human principal defines the boundaries; the agent operates within them. If the agent attempts to exceed its SPT parameters, the transaction is rejected at the protocol level, not after the fact.

What does the early adopter landscape reveal?

The early adopters listed in the MPP payments directory reveal the emerging shape of the agent economy. Browserbase enables agents to pay per browser session. PostalForm allows agents to pay to send physical mail. Prospect Butcher Co. enables agents to order sandwiches for human pickup. These are not demonstration projects - they are live commercial services accepting machine-initiated payments.

The diversity of use cases - from digital infrastructure to physical goods - demonstrates that agent-to-service commerce is not confined to a single vertical. The long tail of the agentic economy is forming across every category of commercial activity.

Browserbase: agents paying per browser session (digital infrastructure)
PostalForm: agents paying to send physical mail (physical services)
Prospect Butcher Co.: agents ordering sandwiches for human pickup (food and beverage)
Over 100 services listed at launch across multiple categories

How does MPP relate to the protocol proliferation problem?

The launch of MPP intensifies the protocol proliferation problem the AXD Institute has been tracking. Merchants must now navigate at least five distinct protocol standards: ACP (Stripe/OpenAI), MPP (Stripe/Tempo), x402 (Coinbase), UCP (Google), and network-specific protocols from Visa and Mastercard.

Stripe's strategic decision to support both MPP and x402 confirms the AXD observation that the agentic payment layer will be multi-protocol, not winner-take-all. The question is not which protocol wins but how the stack converges - and who provides the trust verification layer that sits above all protocols.

What does Visa's endorsement signal?

Visa's decision to support MPP for card-based payments across its global network is a protocol convergence signal. When the world's largest card network endorses an open payment protocol co-authored by a payment processor and an AI infrastructure company, the strategic direction is clear: card networks are positioning themselves as protocol-agnostic trust layers.

Rather than building walled gardens for agentic payments, Visa is becoming the trust verification layer that validates agent identity and authorises transactions regardless of which protocol initiated them. The infrastructure stack is crystallising: protocols at the application layer, card networks at the trust layer, and issuers at the authorisation layer.

AXD Frameworks

Protocol Proliferation Agentic Payments Trust Architecture Machine Customer

Frequently Asked Questions

What is the Machine Payments Protocol (MPP)?

The Machine Payments Protocol (MPP) is an open standard launched by Stripe and Tempo that enables AI agents and digital services to coordinate payments programmatically. It uses a four-step sequence - request, price, pay, deliver - that requires no human interface. MPP introduces Shared Payment Tokens (SPTs) for programmable, scoped authorisations.

How does MPP differ from other agentic payment protocols?

MPP focuses on agent-to-service payments with streaming micropayment capabilities via Tempo's runtime. ACP (Agentic Commerce Protocol) handles broader agent-merchant commerce flows. x402 uses the HTTP 402 status code for chain-agnostic thin-layer payments. UCP (Universal Commerce Protocol) from Google focuses on discovery and full-cycle commerce. Each serves a different layer of the agentic commerce stack.

What is the protocol proliferation problem?

The protocol proliferation problem refers to the growing number of competing standards that merchants must support simultaneously - ACP, MPP, x402, UCP, and network-specific protocols. Rather than a single standard emerging, the agentic commerce infrastructure is forming a multi-protocol stack. Merchants need multi-protocol readiness, and card networks are positioning as protocol-agnostic trust layers above the application protocols.

Can MPP handle both crypto and traditional card payments?

Yes. MPP supports both stablecoins and fiat payments via Shared Payment Tokens. Visa has extended MPP to card-based payments across its global network, and Lightspark has extended it to Bitcoin Lightning. This dual-rail approach - supporting both traditional and crypto payment methods - is the architecture the AXD Institute identified as necessary for the protocol convergence phase.

About the Author

Tony Wood

Founder, AXD Institute

Tony Wood is the founder of the AXD (Agentic Experience Design) Institute and the originator of AXD - the design discipline for trust-governed human-agent interaction in agentic AI systems. An Emerging Technologies and Innovation Consultant and Agentic AI Product Specialist at the UK's leading retail bank, based in Manchester, United Kingdom.

LinkedIn X / Twitter Full Profile