Skip to main content
KYA Agent identity

Identity for the agentic web.

Verifiable identity, capability scope, and accountability for autonomous AI agents acting on behalf of customers and businesses — extending KYC/KYB to the agent layer.

Book a Demo → Read the primer
Trusted by 500+ institutions across 180 jurisdictions
FAB NETS MSIG Moomoo Syfe GLDB
1:1
Action attribution
Every action bound to the agent that performed it and the human who authorised it.
100%
Capability attestation
Signed manifest of model, version, permitted actions, and policy — checked at every call.
Day 1
Regulator-ready
Agent inventory aligned with IMDA AI frameworks — one audit bundle exportable on demand.
How it works

Three things that make agentic AI defensible.

Verifiable agent identity

Cryptographic identity bound to operator, model lineage, and authorising human. No identity, no execution — by construction.

Signed capability manifests

Each agent declares model, version, permitted actions, data scopes. Out-of-scope calls blocked at the boundary — not after the fact.

Immutable agent audit log

Every invocation, attestation, denied action, and policy version on an append-only log. One bundle for supervisory examination.

Recognised by the Industry*
Chartis FCC50 2026 RegTech100 Singapore Top Fintech 2026 Regulatory Leader 2025 ALB Pan Asian 2025 Fintech Frontiers 50 Chartis FCC50 2025 Top 10 Singapore Fastest-Growing 2020 Top 50 High-Growth Asia-Pacific 2020 MAS FinTech Awards
* In collaboration with Cynopsis Solutions
FAQ

Common questions about Know Your Agent

What is Know Your Agent (KYA) and why does it matter?
Autonomous AI agents are increasingly initiating financial actions on behalf of humans and businesses. KYA is the verification layer that establishes (a) the agent's identity, (b) the principal it represents, (c) the capability scope it's authorised for, and (d) the accountable human owner. Without KYA, AI-initiated actions cannot satisfy FATF / FinCEN identification standards.
How does an AI agent get a verifiable identity?
WIDTH issues each agent a cryptographically signed agent credential bound to the principal's verified KYC or KYB envelope. The credential includes the agent ID, model fingerprint, capability manifest, and signing key — verifiable by any counterparty financial institution.
What's a capability manifest?
A signed declaration of what the agent is permitted to do — e.g., 'initiate payments up to S$5,000', 'query account balance', 'open accounts on behalf of principal X'. Counterparty systems can refuse any request that exceeds the manifest.
Which regulatory frameworks does KYA align with?
Singapore IMDA Model AI Governance (agent accountability), MAS FEAT (transparency, fairness, accountability), EU AI Act (high-risk systems), UK FCA AI Update (accountability + governance), and the emerging FATF guidance on AI-mediated transactions.
How is KYA different from KYC and KYB?
KYC verifies a natural person; KYB verifies a legal entity. KYA verifies an autonomous agent and the chain of accountability back to a verified principal (person or company). Think of it as a third primitive that completes the identity layer for the agentic web.
Can KYA work with non-WIDTH agents?
Yes. KYA is built on open primitives (DID + Verifiable Credential) so agents from any framework — LangChain, AutoGen, custom — can be enrolled. WIDTH's role is the verification + revocation registry, not the agent runtime.
Related modules

One platform. Shared identity envelope.

AI · Case Review

AI Reviewer

Hold AI reviewers themselves to a verifiable audit chain.

Learn more →
KYC · Individuals

KYC Onboarding

The human principal that an agent acts on behalf of.

Learn more →
KYB · Corporate

KYB Onboarding

The corporate principal authorising the agent.

Learn more →

Agentic AI, defensibly governed.

30-minute call. Live agent provisioned. Full KYA audit bundle exported at the end.

Book a Demo → Explore AI-Native