Know Your Twin is the verification practice for confirming an AI twin's identity, authority and accountability — the twin-native answer to Know Your Agent (KYA). Built on F-ACT, the open standard for agent conformance and trust.
Know Your Twin, powered by F-ACT.
"Know Your Customer" reshaped finance. "Know Your Agent" is reshaping AI governance. But the age of AI twins — autonomous digital counterparts that act on behalf of people, organisations and even other agents — demands something sharper: a verification practice native to the twin model, not retrofitted from agent-only thinking.
Most AI agents today operate without verifiable identity. They act, delegate, access data — but no one can confirm who authorised them, or revoke that authority in real time.
An AI twin carries a deeper relationship — it represents a person, a company, a role. Its actions carry reputational and legal consequence. Verification must match that weight.
The EU AI Act, NIST AI RMF and emerging frameworks all demand traceable, auditable, governable AI. Know Your Twin maps directly to those requirements through F-ACT.
Know Your Twin maps to the emerging industry "Know Your Agent" (KYA) pillars — grounded in F-ACT's five normative questions (GUARD™) so every pillar is backed by an open, auditable standard.
Who is this twin, and who stands behind it?
Every AI twin must carry a verifiable, human-traceable identity — linked to an accountable principal, not just a model name or an anonymous endpoint.
What is it allowed to do — and what is off-limits?
A twin's scope of authority must be explicit, bounded and machine-readable: what it may access, decide, delegate, and where the hard limits sit.
Can it prove where it came from and what it follows?
The twin must demonstrate adherence to a declared governance framework — its policy version, training provenance, and the standard it conforms to.
Is it still trustworthy right now?
Verification is not a one-time event. A twin's trust status must be monitored, re-evaluated, and capable of degrading or being revoked in real time.
Would the evidence hold up under scrutiny?
Every consequential action must leave a tamper-evident trail — an evidence record that an auditor, regulator or harmed party can inspect independently.
When a twin passes Know Your Twin verification, it earns the Know Your Twin Verified mark — a conformance badge backed by F-ACT evidence, not a self-declaration.
The Know Your Twin Verified badge is designed to be displayed on directory listings, marketplaces and platform registries — a trust signal visible wherever a twin is discovered.
Behind the badge sits a full evidence pack: the twin's identity chain, authority scope, policy version, conformance tier and audit trail — exportable and inspectable.
The mark reflects real-time status. If a twin's authority is revoked, its conformance lapses, or its governance posture degrades, the verified status updates accordingly.
Know Your Twin inherits F-ACT's conformance ladder (L0 Unattested → L1 Declared → L2 Enforced → L3 Provable), so the badge communicates depth, not just presence.
We are entering an era where humans, AI twins and humanoid agents will share the same spaces — digital, physical and hybrid. They will collaborate, transact, compete and coexist.
In that world, the question is no longer "can this model perform?" but "should this twin be trusted here?" Know Your Twin is the verification ritual that makes that question answerable — before access is granted, before authority is delegated, before the twin acts.
The places and spaces where people and intelligent agents meet will need trust infrastructure that works for all of them. Know Your Twin is one piece of that foundation.
Know Your Twin is not a separate standard — it is the branded, twin-native implementation of the F-ACT open standard (Framework for Agent Conformance & Trust). F-ACT provides the five normative questions (GUARD™), the conformance ladder, and the royalty-free pledge that make Know Your Twin verifiable, auditable and open.