Most of your traffic isn't a person any more.
53% of web traffic is now automated. Humans are 47%, and falling. Every system you own was built on the opposite assumption — that there's a person on the other end unless something proves otherwise. That assumption is now wrong more often than it's right, and everything downstream of it is quietly mispriced: your signups, your reviews, your ad impressions, your fraud models.
Know Your Human fixes the assumption. A real, ID-verified human verifies once, and you get a signed receipt that anyone can check — offline, with no call to us, using any standard library.
Web Bot Auth, Know Your Agent and the Trusted Agent Protocol all reference a verified human principal. None of them provide one.
That missing layer is what this is. Every standard being written for agents assumes someone, somewhere, established that a real person authorised the action — and then leaves that step to you. KYH is that step, as a credential you can hand to anyone. How it fits alongside Know Your Agent →
Why now
Agents can act for us
World shipped AgentKit (June 2026) to tie an AI agent to a person — the market now agrees a real human has to be behind the agent.
Money moved on-chain
The GENIUS Act's stablecoin framework finalizes in 2026 and Open USD (OUSD) launched across a Visa / Mastercard / Stripe / Coinbase / BlackRock consortium — value can now reach an agent in seconds.
The gap is the human
Roughly $94.9B a year is lost to gaps in digital identity. Faster rails with no human layer just move fraud faster.
KYC vs. KYH
Know Your Customer answers who. Know Your Human answers whether you're a real human at all — and it's built to be reused, by you and by the agents acting for you.
| KYC — Know Your Customer | KYH — Know Your Human | |
|---|---|---|
| Question answered | Who is this legal identity? | Is this a real, ID-verified, live human? |
| Reusable across sites | No — repeated per institution | Yes — verify once, prove everywhere |
| Speed to re-verify | Minutes — re-scan / re-upload | Under 1ms — a credential check |
| Privacy | Exposes identity documents | Biometrics stay on device; only a signed attestation leaves |
| Works for AI agents | Not designed for it | Yes — binds human + agent + payment |
| Independently verifiable | Trust the vendor | Public JWKS — anyone can check |
KYH complements KYC — it doesn't replace compliance. Many flows want both: KYC for the legal identity, KYH for the living human (and their agent) behind it.
The moat: verify once, prove everywhere
Speed
The industry re-verifies the same human on every site. KYH is verify-once: after that, proving humanness is a sub-millisecond credential check — no repeat face scan, no new upload, no queue.
Privacy
Liveness and attention run on the user's own device. There's no central biometric database and no orb — only a signed attestation ever leaves the phone.
Portability
The proof is a plain ES256 credential. Any website, payment rail, or agent framework can verify it against public JWKS — no SDK lock-in, no walled garden to enroll in first.
How it works
Identity + on-device liveness and attention. Biometrics never leave the device.
5arz issues a signed Proof-of-Human credential — an ES256 JWT bound to the person (and, optionally, their agent + payment).
Show the credential to any site, rail, or agent. No re-scan, no re-upload.
They validate it against 5arz's public JWKS with any JWT library. No 5arz SDK required.
Where Know Your Human is needed
Agentic commerce is the first market to demand it — but every channel AI can flood needs proof of the real human. One credential standard covers them all.
Agentic commerce
Prove a real, accountable human is behind the AI agent making the purchase. See agentic commerce →
AI training data
Signed provenance that a real human produced each label, rating, or judgment — clean human data AI teams can trust. Proof of Work →
Content authenticity
A verified-human mark for reviews, posts, and creative work as AI-generated content floods every feed.
Trust & safety
ID-verified, live humans for marketplaces, dating, and social — without doxxing anyone. Proof of Personhood →
Verified-human ads
Prove a real, attentive human — not a bot — saw the impression. Verified human traffic →
DeFi / RWA
Bind a verified-human credential to a wallet, then read isVerifiedHuman(wallet) on-chain for gating on wallets bound to ID-verified humans.
The human layer for GENIUS-Act rails and OUSD
The U.S. GENIUS Act put a federal framework around payment stablecoins — with rules finalizing in 2026 — and the Open Standard consortium (Visa, Mastercard, BlackRock, Coinbase, Stripe, BNY + 140 institutions) launched Open USD (OUSD). Money can now move to an AI agent in seconds. What those rails still can't tell you is whether a real, accountable human is on the other side.
Human proof on the rail
5arz binds a verified human to any stablecoin transaction — pass a paymentRef from OUSD, USDC, Agent Pay, or Stripe and get back a signed credential clipping human + agent + payment together.
Compliance-friendly by design
Proof of a real, live human — with biometrics kept on-device — supports the anti-fraud and accountability posture regulated issuers need, without turning into another identity honeypot.
Rail- and issuer-agnostic
The credential binds to any settlement token via a payment reference. As OUSD and the GENIUS framework go live, 5arz is the portable human layer on top.
FAQ
What is Know Your Human (KYH)?
A reusable proof that a real, ID-verified, live human is behind an action. KYC proves a legal identity for compliance; KYH proves humanness itself — that this is a genuine person, not a bot or an unattended AI agent. Verify once with 5arz and hold a signed credential any site, rail, or agent can check in under a millisecond via public JWKS.
How is KYH different from KYC?
KYC answers "who are you" per institution, for compliance — slow, repeated, and identity-exposing. KYH answers "are you a real human at all," is reusable across sites and by your AI agents, and keeps biometrics on-device. They pair well: KYC for legal identity, KYH for the living human behind it.
Does KYH work for AI agents?
Yes. An agent can present proof that a real, accountable human authorized it. 5arz binds the human, the agent, and the payment into one signed credential, so the agent proves it represents a genuine person instead of running anonymously.
What makes it reusable and fast?
You verify once and hold a portable credential. Presenting it later is a cryptographic check against 5arz's public JWKS — no repeat face scan or upload. Verification takes under one millisecond with any standard JWT library.
Is my biometric data stored?
No. Liveness and attention run on your own device; only a signed attestation leaves it. No central biometric database, no orb.
Trust no one. Verify anyone.
A KYH credential is a standard ES256 JWT. Validate it against 5arz's public keys — no 5arz SDK, no account, no trust in us required. It's an open standard, not a walled garden.
Verify it yourself
GET api.5arz.com/.well-known/jwks.json
→ verify(credential, jwks)
Any standard JWT library. A sub-millisecond cryptographic check that runs locally and offline.
The open standard
A published spec — ES256 / JWKS plus a W3C Verifiable Credential profile. Read it, build on it, verify against it. See the standard →
Backed by property law
Grounded in UCC Article 12 (Controllable Electronic Records) and patent-pending (US 63/979,052 + 63/979,474) — a proof-of-human layer with a real legal class underneath.
Works with the rails: USDC and Stripe today, Mastercard Agent Pay, and OUSD-ready as Open USD goes live — plus native agent integration via a live MCP server and OpenAPI 3.1.
Know the human. Not just the customer.
Start verifying in about thirty seconds.
Unlimited human verification, free forever. Two fields, no card, no sales call — or let your agent register itself through MCP.