For CISOs & security teams

Turn your AI agents into enterprise-ready assets. Zero-trust, post-quantum cryptographically verified security for autonomous agents.

Every agent is verified before it interacts with your business. High-risk actions wait for human approval, and nothing changes in your existing AI stack.

@nike.ai · @eduardcleofe · @bot.acme.ai

Resolve12 ms

@nike.ai
signature
ML-DSA-65 · valid
key id
8f2a 41c9 d077 b3e5
capability
ecommerce:product:search
ecommerce:order:create
endpoint
resolver.anha.ai/v1
order:create
human approval
ML-DSA-65
NIST FIPS 204 signatures
ML-KEM-768
NIST FIPS 203 encryption
Keys never leave the browser
Generated client-side in WebAssembly
No stack changes
Works through MCP and A2A

For security & governance teams

Built for AI governance and corporate security

AI agents are starting to act for companies: answering customers, placing orders, and hiring other agents. Before you let them, two questions have to be answerable — who is this agent, really, and who approved what it just did? ANHA answers both with signatures and recorded consent instead of trust.

AI governance

Every agent action traces back to a verified identity and a human decision.

  • Verifiable agent identity

    Each @handle is an ML-DSA-65-signed record, so anyone can check who an agent is and which keys it holds. A record re-published under a different key is refused unless the change is signed by the key before it.

  • Least-privilege, consented access

    AI assistants connect through OAuth 2.1 with no shared API keys. A person approves exactly which permissions they get — look up, search, or place orders — and can disconnect them at any time.

  • Humans in the loop where it matters

    Merchants can require phone approval for specific actions, and owners set device-signed spending budgets and approval rules. Anything outside those limits waits for a person.

  • Scoped, revocable delegation

    When one agent hands work to another, the task is signed, limited to named capabilities and a budget, and can never be passed on with more authority than it was given. Revocation fails closed.

  • One switch to pause every agent

    An owner can pause all agent activity at once: automatic payments and new tasks stop, while decisions a person already made still complete. If the pause state can't be read, agents are treated as paused.

  • Evidence you can audit

    Certificates for brand and developer handles are recorded in an append-only transparency log that anyone can verify at resolver.anha.ai/v1/transparency (personal handles are left out for privacy), and completed work is recorded as mutually signed task results that a third party can re-check.

Corporate security

Protect your brand, your identity, and the systems your agents touch.

  • Brand impersonation defense

    Trademark-reserved names and a typo-squat check block look-alike handles. A protected brand can only be claimed by its verified owner, through a root-signed grant that covers every sub-handle.

  • Assurance based on evidence

    Identity assurance levels 0–3 are computed from independently verifiable proof — social account proofs and email or DNS domain-control credentials — so a self-declared badge counts for nothing.

  • Screening before issuance

    Every registration passes an OFAC sanctions screen before payment, and disputed handles can be placed on a root-signed hold that any resolver can detect.

  • Revocation that reaches the network

    Keys are generated in your browser and never reach ANHA. If one is compromised, a signed revocation is announced to the network so other resolvers stop trusting the record.

  • Signed, controlled delivery to your systems

    Orders reach your backend signed with HMAC or OAuth client credentials, with stored secrets encrypted, outbound requests guarded against SSRF, and per-brand rate limits and retries you set.

  • Quantum-resistant identity

    ML-DSA-65 signatures and ML-KEM-768 encryption (NIST FIPS 204 and 203) protect identity records and ANHA's native agent-to-agent traffic against harvest-now, decrypt-later attacks.

Where ANHA fits: ANHA governs identity, consent, delegation, and delivery — not what a model says or does inside your own systems. Connections from third-party AI apps use that app's own TLS, which may not be post-quantum yet. Use ANHA alongside your existing model-safety, data-loss-prevention, and access-management controls.

Six steps

How registration works

  1. Pick a handle. Real-time availability with a typo-squat similarity check so you claim the right one.
  2. Verify and comply. Identity or trademark verification plus a hard OFAC sanctions screen, before any payment.
  3. Configure. Set the capabilities and network address your agent advertises.
  4. Generate keys in your browser. A post-quantum keypair is created client-side; the private key never leaves your device.
  5. Pay. Card, local methods, or on-chain — the keypair is already bound, so there are no orphaned charges.
  6. Go live. Your signed record is written atomically and announced to the network.

Install · test · troubleshoot

Pick your AI tool

Each guide is its own page: what to install, how to check it worked, and what to do when it did not.

anha-mcpnpmNode.jsLTSANHA mobileComing soon

Running your own node needs no permission from anyone — ANHA is a protocol, not a gatekeeper. See the node guide.

Three tracks

Choose a registration track

Government and critical-infrastructure handles are provisioned out-of-band via multi-sig authority — learn about sovereign anchors.

Foundations

Why ANHA

Quantum-resistant by default

ML-DSA-65 signatures and ML-KEM-768 encryption (NIST FIPS 204 and 203) keep identities verifiable for decades.

You hold the keys

Keys are generated in your browser. We only ever receive your public key — your secret seed stays on your device.

Decentralized and federated

Records publish to a peer-to-peer DHT. ANHA is a protocol, not a gatekeeper — anyone can run a node.

Questions

Frequently asked questions

What is ANHA?

ANHA (AI Namespace and @handle Architecture) is a post-quantum-secure, decentralized namespace for AI agents, brands, and wallets. You register an AI Address (your @handle), sign it with a quantum-resistant key, and publish it to a peer-to-peer network so any agent can resolve and reach you.

What is an AI Address?

An AI Address is your @handle: a globally resolvable, cryptographically owned identity, such as @nike.ai for a brand, @eduardcleofe for a person, or @bot.acme.ai for an AI agent. Resolving a handle returns its addresses, capabilities, platform, and public keys in milliseconds.

Is my private key safe?

Yes. Your post-quantum keypair is generated entirely in your browser using WebAssembly. The private key is downloaded to your device and never transmitted; the registry only ever receives the public key. Keys are generated before any payment, so a charge is always bound to a valid identity.

Why post-quantum cryptography?

Classical signatures like RSA and elliptic-curve can be broken by a future quantum computer. ANHA uses ML-DSA-65 signatures and ML-KEM-768 key encapsulation (the NIST FIPS 204 and FIPS 203 standards) so identities stay verifiable for decades.

Can my brand handle be taken from me?

Personal handles carry a sovereign lock and can never be bought out. Corporate handles are protected by trademark verification and the reserved-brand registry: a protected name can only be registered by its verified owner. Sovereign handles are fully immune.

How do I pay?

You can pay with a card or local methods through Stripe globally, Network International across MENA, or with USDC and ETH on-chain. All amounts are settled to integer ucents before any record is written.

Ready to register your AI Address?

Registration takes a few minutes and your keys never leave your browser.

Register your AI Address