Agentic Experience Design (AXD) is the discipline for designing trust-governed relationships between humans and autonomous AI systems. Founded in September 2024 by Tony Wood in Manchester, United Kingdom, AXD addresses how humans delegate, calibrate, observe, interrupt, and recover trust in agentic AI.
| 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 legibility is the design requirement that an autonomous AI agent's identity, actions, reasoning, capabilities, and limitations must be readable by humans, institutions, and other agents. It operates at three levels: identity legibility (who authorised this agent?), action legibility (what has it done and why?), and capability legibility (what can and can't it do?). Legibility is the prerequisite for trust in agentic systems.
AI explainability focuses on why a model made a specific prediction or decision. Agent legibility is broader - it encompasses the entire agent: its identity and authority chain, its complete action history, its operational constraints, and its known limitations. An agent can be explainable (its model is interpretable) without being legible (its overall behaviour and authority are not readable). Legibility is the design requirement; explainability is one component of it.
Machine legibility is the requirement that agents must be readable by other agents, not just by humans. In multi-agent systems, agents must verify each other's credentials, evaluate track records, and understand capabilities through structured, machine-readable formats. Machine legibility enables agent-to-agent trust without human intermediation. Infrastructure like Agent Registries and the Universal Commerce Protocol (UCP) address machine legibility at scale.
Trust cannot be formed in an agent whose actions are opaque. Agent legibility enables the three components of trust architecture: competence trust (believing the agent can do what it claims), integrity trust (believing it will stay within scope), and benevolence trust (believing it acts in the human's interest). Without legibility, trust calibration is impossible - humans will either over-trust or under-trust their agents.
Design for legibility at three layers: identity (verifiable credentials, delegation chains, Agent Registry registration), action (audit trails, decision logs, reasoning traces, post-hoc explanations), and capability (operational envelope documentation, constraint specifications, limitation disclosures). The AXD Practice provides frameworks for each: the Explainability Standard, the Onboarding Framework, and the Multi-Agent Orchestration Visibility Model.
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.