Agent-to-Agent Commerce — an AXD Institute resource on agentic experience design, agentic commerce, trust architecture, and human agent interaction. Founded by Tony Wood..
| Dimension | Traditional UX | Agentic Experience Design (AXD) |
|---|---|---|
| Primary material | Attention and affordance | Trust and delegation |
| User state | Present, navigating | Absent, delegating |
| Design output | Screens and interfaces | Outcomes and constraints |
| Temporal model | Session-based | Relationship-based |
| Success metric | Task completion | Trust calibration |
Agent-to-agent commerce is the practice of AI agents transacting directly with other AI agents - a buyer's procurement agent negotiating with a supplier's sales agent, without human involvement in the transaction itself. It works through structured protocols: discovery (agents find each other through capability advertisement), negotiation (agents exchange structured proposals within their delegated authority), settlement (agents execute agreed transactions through atomic protocols), and verifica
Agent-to-agent trust establishment follows a structured protocol defined by the AXD Trust Architecture framework. Identity verification: each agent presents verifiable credentials proving its identity and the identity of its principal organisation. Authority verification: each agent provides evidence of its delegated authority - what it is authorised to negotiate, commit to, and execute. History exchange: agents share or reference their transaction history, reliability metrics, and trust scores
Designing negotiation protocols for AI agents requires structured multi-attribute frameworks. Define the negotiable attributes (price, quantity, delivery, quality, payment terms, warranties) and their acceptable ranges for each party. Implement proposal-counterproposal exchange protocols where each offer is a structured document with explicit values for every attribute. Build concession strategy frameworks that enable agents to make rational trade-offs within their delegated authority. Include m
Human oversight for agent-to-agent commerce operates at three levels. Authority setting: humans define the agent's mandate - what it can negotiate, what limits apply, and what terms are non-negotiable. This happens before the agent engages in commerce. Exception handling: humans intervene when the agent encounters situations outside its mandate - new counterparties, unusual terms, authority boundary approaches, or negotiation deadlocks. This happens during commerce when the agent escalates. Post
Dispute handling in agent-to-agent commerce requires protocol-defined resolution mechanisms because human negotiation is not the default. When fulfilment does not match agreement, the buyer agent raises a structured dispute containing: the specific terms that were not met, the evidence of non-conformance, and the requested remedy. The supplier agent responds with a structured defence or acknowledgement. If agents cannot resolve the dispute through protocol exchange, the system escalates through
Agentic Experience Design (AXD) is a new discipline for the age of autonomous AI. It addresses trust architecture, delegation design, and human agent interaction — the core challenges of agentic commerce and agentic shopping.