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 |
Agentic KYC is the extension of Know Your Customer (KYC) compliance frameworks to accommodate autonomous AI agents that transact on behalf of human principals. It adds three verification layers beyond traditional KYC: agent identity verification (authenticating the agent itself), delegation chain verification (tracing authority from agent back to human authorisation), and behavioural monitoring (ensuring the agent operates within its delegated authority scope).
Traditional KYC verifies the identity and risk profile of human customers or legal entities. Agentic KYC must verify not only the human principal but also the autonomous agent's identity, the scope of its delegated authority, the integrity of the delegation chain, and the agent's ongoing behavioural compliance. Traditional KYC asks 'who is this customer?' Agentic KYC adds 'what is this agent authorised to do?' and 'is it operating within those boundaries?'
Know Your Agent (KYA) is the AXD Institute's principle that institutions must verify the identity and authority of autonomous agents with the same rigour applied to human customers. KYA encompasses agent identity verification (cryptographic credentials), principal verification (confirming the human behind the agent), authority scope verification (what the agent is authorised to do), and continuous behavioural monitoring (ensuring ongoing compliance with delegated authority).
Delegation chain verification is the process of tracing an autonomous agent's authority back through every link in the chain to the original human or organisational authorisation. In complex B2B environments, delegation chains can span multiple levels. Every link must be verified - if any delegation has been revoked, exceeded, or expired, the entire chain is invalid. Delegation chains must be machine-readable, cryptographically signed, and verifiable in real time.
No major jurisdiction has yet published comprehensive regulations specifically for autonomous agent KYC. However, regulators including the UK FCA, EU EBA, and US FinCEN are monitoring autonomous agents in financial services. The EU AI Act includes provisions for high-risk AI systems in financial services. The AXD Institute recommends that institutions develop agentic KYC frameworks now - before regulation mandates them - to be better positioned when requirements are formalised.
Know Your Customer (KYC) is the regulatory foundation of financial services - the set of processes by which institutions verify the identity, assess the risk, and monitor the activity of their customers. KYC was designed for a world in which every customer is a human being. That world is ending. When an autonomous AI agent approaches a bank to open an account, initiate a payment, apply for credit, or execute a trade, the existing KYC framework breaks down at every level. The agent is not a natural person - it has no passport, no address, no date of birth. The agent is not a legal entity - it has no articles of incorporation, no registered directors, no beneficial ownership structure. The agent exists in a regulatory gap between natural persons and legal entities, and current KYC frameworks have no category for it. This is not a theoretical problem. Autonomous trading agents already operate in financial markets. AI-powered payment orchestration systems process transactions without human involvement. Procurement agents negotiate and execute contracts on behalf of enterprises. These Every autonomous agent must have a verifiable identity - not a human identity, but a machine identity that can be authenticated, tracked, and audited. This requires agent identity standards: cryptographic credentials that prove the agent is what it claims to be, issued by a recognised authority, and revocable if the agent is compromised. The agent's identity must be distinct from its principal's identity - the bank must know both who the agent is and who it represents. Agentic KYC does not replace traditional KYC - it extends it. The human or organisation behind the agent must still pass standard KYC checks. But agentic KYC adds a critical question: has this verified principal actually authorised this specific agent to act on their behalf? The delegation must be verifiable, not assumed. A stolen or forged delegation credential is the agent equivalent of identity fraud. Traditional KYC asks: